Document G5Ld0zr24ZradOLpayVe9YNXv

Respiratory Health in Workers Exposed to Man-made Vitreous Fibers1-3 /tsV f ` HANS WEILL, JANET M. HUGHES, YEHIA Y. HAMMAD, HENRY W. GLINDMEYER, III, GIL SHARON, and ROBERT N. JONES Introduction Man-made vitreous fibers (MMVF), which include fibrous glass, mineral wool (rock and slag), and ceramic fi bers, are manufactured from glass, from natural rocks and minerals, or from slags, the residues of smelting. They are used extensively in thermal, acoustical, and electrical insulation products as well as in reinforcing plas tics and resins. With many millions of tons of prod ucts with MMVF being manufactured annually in the United States and Eu rope, thousands of workers are ex posed to these fibers in the workplace. The U.S. National Institute for Occu pational Safety and Health (NIOSH) estimated that 200,000 U.S. workers experienced occupational exposure to fibrous glass in the early 1970s (1). Since that time the industry has had ad ditional growth, some of it because of the substitution of these fibers for asbestos. The most commonly recognized ef fects of contact with MMVF are acute irritation of the skin and, to a lesser ex tent, the eyes and upper respiratory tract, with symptoms abating after ces sation of exposure. Although acute ef fects on the lungs have not been ob served, the fibrous nature of these sub stances has resulted in concern about possible long-term respiratory effects. Moreover, at a time when there is an evolving consensus that biological ef fects of fiber exposure are related to di mension (with long, thin fibers appar ently posing the greatest risk), environ mental studies have shown that airborne dusts in all plants manufacturing MMVF contain fibers having a wide range of diameters and lengths, a high proportion being respirable. The rela tive frequency of fine fibers depends on the nominal diameter of the fiber being manufactured, with production of (nominally) small-diameter fibers resulting in the greatest airborne con SUMMARY A study of tho respiratory hoatlh of wortore who product mtrvmado vitreous flbore (MMVF) Includ'd 1.02* man (nredUn iangth of amptoymant, IS yr) tn T fibrous glass and mlrrecrt woof plants. Htalth aaaaaamant was by respiratory quastlonnalra, ILO I960 Classification chast radtogrsph readings, and pulmonary function tasting. Rtsulls from an aarllar anvlronmantal surrey ware combtnad with |ob hlatortas to dartva Individual aipoture aatlmatas. Tha study popu lation was found to ba ganarslly haalthy, with rasptrstory symptoms not relatad to Iha Hbar as- posura and no datad ad advarea lung function conaaguancas of that axposure. Radiograph readings revsslad low catagory prolusion of small opadllaa: of *A1, A3 (7%) In catagory (VI, 25 CO (3%) In 1/0. and I In 1/1, with nona In highar catagortas. Prevalanca of small opacitlas Incraasad --i with aga and clgaretta smoking. A greatar pravalanca of small opacitlas wts found among cur- CD rant mok*re in tha two plant* producing both ordinary and flna llbart (for profusion > 1/0.15% In t--i thasa plants, 2% in all otharsf. Among si smokars and navar smokare, greatar pravalanca was ob> t --, sarvad In only ona of thasa plants (14% varsus 1%), tha plant having Iha hlghar airboma concarv (rations ol flbars tass than 1 In dlamalar. Among currant smokare, pravalanca of small opaci- t--^ Hat of profusion > 0/1 Incraasad with langthaning smploymant {bayond aga and pack-ysare of ___ smoking) In both plants producing ordinary and lino flbare and tha flva othar plants comblnad. It ___ la conclud'd that axposure to MMVF with small diamatare may laad to low-iaval prolusion of small opacitlas. Howsvsr, without high-laval profusion of thasa opacitlas In a population with a conaldarabla rang# ol axposure durations, a dilfusa tlssua reaction (a.g., fibrosis) saama unilkaiy but cannot batxcludsd am aev aesfi* ore i*U: i:iw-na centrations of long, thin fibers (2). Manufacture of fine MMVF accounts for less than \ V of production in the United States; total production, how ever, continues to increase in response to demand for special purpose insula tion, primarily for aircraft and aero space applications. Although there are certain similari ties between MMVF and asbestos fi bers, there are important differences. Asbestos fibers split longitudinally, creating even thinner fibers, whereas MMVF break transversely, thus pro ducing shorter and^therefore, possibly less hazardous fibers. Overall exposure levels also differ, with measured levels of airborne fiber concentrations in plants manufacturing MMVF being very low compared with those known to occur in occupational exposures to asbestos. In order to assess the respiratory health of workers in this industry, a cross-sectional survey of current em ployees was begun in 1979. This study, in 7 plants manufacturing MMVF be longing to the sponsoring organiza tion, the Thermal Insulation Manufac turers' Association (TIMA), was de signed to include long-term employees exposed to various dimensions of fi brous glass and mineral wool. The study did not include -workers manu facturing ceramic fibers or products containing those fibers. Methods Study Design and Plant Selection In order to have the largest amount of ex posure data available to investigate possible exposure-response relationships, only the 16 plants surveyed in an earlier environ mental characterization study were consid- (Received in original form November 16, I9S2 and in revisedform March 9, I98J) ' From the Tulane University School of Medi cine, Department of Medicine, Pulmonary Dis eases Section, New Orleans. Louisiana. 1 Sponsored by the Thermal Insulation Manu facturers' Association and supported in part by SCOR Grant No. HL-15092-12 from the Naiional Heart, Lung and Blood Institute. 1 Requests for reprints should be addressed to Hans Weill, M.D . Tulane University School of Medicine. Pulmonary Diseases Section, 1200 Perdido Street, New Orleans, LA 70112. 104 MSPiUATODY H ULTH IN MANMADE YTTNEOUS ElEA WOMENS 10$ ' S l 0 0060 I 5 ered for inclusion in the present study (2); categories for the concentration of airborne that study had included the 16 oldest U.S. fibers, as determined by both optical mi plants engaged in MMVF production. croscopy (OM) and electron microscopy Eight plants were excluded for various (EM): 0.001 to 0.01,0.01 toO.l.O.I to 1.0, reasons: textile production, unavailability and greater than 1.0 fibers/ml. Areas were because plant was sold, past asbestos use, classified into these categories based on the and very large size with complex exposures median measurement for that area; the requiring separate study. Seven of the re level assigned to an area was the geometric maining 8 were selected by the investigators mean of that category: 0.0032, 0.032, 0.32. for inclusion: two fibrous glass plants or 1.5 fibers/ml, respectively (6). For sev (Plants I, 2) producing ordinary fibers eral production lines, median concentration (greater than 3*im in diameter), 2 (Plants 3, levels differed from the beginning of the 4) producing both ordinary and fine fibers line to the end, but all jobs on the lines were (1 to 3 pm), I (Plant 3) producing only very labeled identically in the work histories; in fine fibers (less than I pm), and 2 mineral such cases, the median concentration for wool plants (fibers greater than 3 pm). the entire line was "used. Products of these plants were primarily These exposure estimates were used in insulation materials but also included such conjunction with individual job histories to items as roofing supplies, acoustical tiles calculate total accumulated exposure and and panels, automotive liners, and flexible average concentration of exposure for each and rigid fibrous glass ducts. eligible employee. Calculations were per formed separately for fibers identified by Exposure Estimation OM (I to 3 pm in diameter) and by EM (less than I pm), as well as their sums (< 3 pm). The glass commonly used for the manufac Total length of plant employment and time ture of fibrous glass is a borosilicate type: spent in jobs at or above several fiber con the batch usually contains silica, soda ash, centration levels were also calculated for boric oxide, fluorspar, rutile, and clay. The both the OM-sized and EM-sized fibers. materials are conveyed pneumatically from storage silos to a mixer, and then are fed into a furnace. Insulation wool is produced Employee Eligibility by feeding the molten glass through a fore- In order to maximize the likelihood of de- hearth to centrifugal spinners where the fi , tecting adverse health effects resulting from ber is forced out of small orifices in the fiber exposure, a minimum of 10 years of spinner head. The fibers are attenuated by employment was required for participants compressed air or steam, and the binder is from Plants 1 to 4. Because Plants 5 to 7 sprayed on the fibers. The fine and very were small, a minimum of only 5 yr was re fine fibers are produced by passing the mol quired in these plants. Male employees with ten glass through a bushing plate located at 1 month or more in the batch or binder the bottom of a forehearth where large di areas, as well as female employees (less ameter fibers are formed. The fine fibers than 51 of the work force), were excluded. are produced from the primary fibers by Because the study was designed to include flame attenuation. approximately 200 workers from each In slag wool manufacturing, slag and plant, additional inclusion restrictions were coke are melted and allowed to flow in imposed at the larger plants; these excluded open channels to rotary spinners. Fiber at employees with lengthy work in clerical tenuation is achieved by blowing air under positions. In 6 plants, recruitment efforts high pressure around the spinners. were directed at all employees meeting the Because of free crystalline silica in the inclusion criteria; in the largest plant (Plant batch mixing areas and phenol-formalde I), a random sample from the eligible em hyde in the binder mixing areas, workers in ployees was selected for recruitment. those areas were excluded from the study. Using official employee rosters from Estimation of workers' exposure to air each plant (e.g., state tax and social securi borne fibers was based on a survey con ty reporting forms), job histories were ob ducted by the University of Pittsburgh dur tained for 375 randomly selected employees ing 1974 and 1975, in which approximately not originally identified by plant personnel 600 samples were collected in the 7 plants as eligible. Of these, only 4 (1%) qualified included in the present study (2). A similar for inclusion in the study. Thus, identifica survey, using the same sampling methods tion of the target population was excellent, and sample analyses, collected 201 samples with few eligible employees having been in 1980 as part of the present investigation; omitted from recruitment efforts. results showed that exposures had not changed significantly since the earlier study. The Pittsburgh survey provided esti Respiratory Health Assessment mates of fiber concentration levels in all A modified version of the American areas of each plant for all time periods since Thoracic Society-Division of Lung Diseases opening of the plants; these were the same (ATS-DLD) questionnaire (7) was adminis as those used by Enterline and Henderson tered in all 7 plants by the same trained (3) and Enterline and Marsh (4, 5). These interviewer. estimates were obtained by considering 4 Participants were judged to have lower respiratory symptoms if they reported hav ing usual cough, phlegm, or wheezing, cur rent attacks of shortness of breath with wheezing, or breathlessness when walking with others; upper respiratory symptoms if current sinus trouble or postnasal drip; chronic bronchitis if usual cough and phlegm for more than 3 months in each of the last 2 yr. Dyspnea levels were defined as: Grade I. shortness of breath when hurrying on level ground or walking up a slight hill; Grade 2, when walking with others one's own age; Grade 3, when walk ing at own pace on level ground; Grade 4, when washing or dressing. Chest roentgenograms were taken at local facilities and sent to our unit. After all films were received they were placed into random order and read independently by 3 outside readers experienced in the use of the ILO 1980 Classification of Radiographs (8). Films judged unreadable by 2 of the 3 readers were excluded from analysis. The median of the 3 readings was used as the summary reading. For shape of small opacity (rounded or irregular), at least 2 readers had to agree on presence and shape in order for a summary reading to be determined. Pulmonary function testing was con ducted at each plant site in a mobile labora tory using a Pulmolab (Model 5000; Cardio pulmonary Instruments, Houston, TX). This unit measures forced expiratory flows and volumes, lung volumes, and single breath carbon monoxide diffusing capacity (Dlco)- Calibration for all measured pa rameters was performed at the beginning and middle of each testing day. Spirometry was conducted standing; noseclips were used. At least 3 satisfactory forced expiratory maneuvers were re corded. A forced expiratory maneuver was considered satisfactory if its curve had a sharp initial flow and showed smooth con tinuous effort extending until flow ceased or 7 s elapsed. At least 2 curves of satisfac tory shape with forced vital capacity (FVC) within 3V of the largest were required. The values of FVC, forced expiratory flow be tween 25 and 75^i of FVC (FEFn-n). 4d forced expiratory volume in one second (FEV,) used in statistical analyses were averages from the 2 satisfactory maneuvers with the greatest total of percent of pre dicted for FEV,.i and FEF)>.?>. Both FEV,.) and FEV| were computed using backward extrapolation from the steepest part of the curve. Residual volume (RV), measured by the nitrogen washout technique, was consid ered satisfactory when 2 tests were within IO^i or 200 ml. whichever was larger. The smaller RV was used in analysis. Tola! lung capacity (TLC) was calculated as the sum of RV and the larger of FVC or a slow vital capacity. The Dlco was taken as the aver age of 2 measurements within I0*?t of each other in which the alveolar volume (in spired volume plus multibreath residual IOC WilLL. MUOHli. HAUIUO, OLINOMCTCK. SHARON, AND JONH table 1 STUDY PARTICIPANTS BY PLANT Plant" (Fiber Type) 1 (Ordinary) 2 (Ordinary) 3 (Ordlnary/Flne) 4 (Ordlnary/Flne) 5 (Very Fin#) 9 (Mineral Woof) 7 (Mineral Wool) TolA< Nun> bar 2fl0 244 110 217 107 34 SB 1,028 M#dl#n Aq# 48 (29-86)7 44 (28 - 66) 82 (32-84) 48 (30-66) 34 (24-65) 38 (27-80) 43 (24 -6J) 46 (24 - 86) Martian Year# Employed In Pltnl 20 (l0-34)t 17 (10-34) 26 (9-38) 18 (11-31) 9 (4-30) 16 (8-29) 14 (5-42) 16 (4-42) Btiek (%) 14 25 1 8 3 31 0 12 Currant Cigarette Smoker# (%) 47 52 37 36 59 56 59 47 * TTwt* plant* **f* numMrtd. r*p*ct,,iy. 6. 16. 4. i. 11,, H, s. in 1 Rang* na.ionm*nt HixJy (2). Navar Clgaralla Smokara (%) 22 20 25 28 23 35 20 24 volume) (ALVOL) was at least 85% of TLC and the breath-holding time was 10 to M s. Statistical Analysis Muhiple logistic regressions were used to test for a relationship between possible ex planatory variables and dichotomous re sponse variables (e g., presence of small opacities of a specified profusion level, presence of symptoms); multiple linear re gressions were used in the case of contin uous response variables such as the pulmo nary function measurements. These regres sions were run on the smoking groups separately as well as on these groups com bined, allowing for smoking main effects and interaction terms. Tests for the signifi cance of the exposure variables were per formed after allowing for other factors such as age, height, race, and smoking. Response variables were examined for in fluence of all exposure variables; fiber type, plant, average and cumulative dust levels (by both optical and electron microscopy counts, as well as their sum), length of exposure, and time spent at or above spe cific fiber concentration levels. Logarithms of the exposure measurements were also used to allow for a possible exponential relationship. Results Participation Of the 1,148 employees asked, 1,028 (90%) participated in some phase of the study; participation rates were sim ilar in all 7 plants. Health response data were, for various reasons, not complete for all participants; of those eligible, interviews were obtained for 79%, function testing was done in 78%, and radiography was done in 83%. Of those who did not participate on-site, most (67%) were unavailable at the time of testing (e.g., vacation, sick leave). Of the 1,028 participants, 35 had been employed in a plant job with silica iexposure, and an additional 36 may have had silica exposure in mainten ance jobs. Analyses comparing these 2 groups of workers with those with no silica exposure found no differences in any respiratory health measures and these workers were retained in the study. Population Description Median age of the 1,028 participants was 46.3 yr; 12% were black and 88% were white. Median length of employ ment in the plants was 18.0 yr; 21 par ticipants were found not to have the minimal required length of employ ment because of leaves of absence but were retained in the study. Among the eligible employees, those who participated in various study phases and those who did not had com parable ages and employment dura tions. For example, the radiography participants had mean age of 45.9 yr and mean employment of 19.0 yr, whereas these parameters for nonpar ticipants were 46.6 yr and 18.9 yr, respectively. Interviews with 912 employees found 48% cigarette smokers (currently smoking or stopped within the past year), 29% ex-smokers (stopped more than 1 yr before), and 24% never cig arette smokers. Of these 912 employees, 187 (21%) had been employed previ ously in jobs potentially harmful to re spiratory health (27 of ihesc 187 had multiple previous exposures); 38 had worked for more than 1 yr in a foun dry, 29 for more than 1 yr in mining or quarrying, 120 had sandblasted, 12 had worked with asbestos, and 15 had worked in a cotton mill. There were considerable differences among the plants in age of workers and length of employment (table 1). There was also a wide range in the percent ages reporting previous employment in TABLE 2 PREVALENCE f/V) OF SELECTED SYMPTOMS BY SMOKING CATEGORY Symptom Cigaratla Smoking Category Never amoker, n = 215 Ea-smoker, n 3= 261 Curran! smoker, n = 436 Total, n = 912 Lower Reap!raiory 22 22 56 38` Upper Raaptraiory 37 39 42 40 Chronic Brorv chill* Dyapnaa (Qrada D Dyspnaa (Grade > 2) Shorlneae ot Braath with Wheezing Wheezing 5 24 2 35 16 42 9` 36` 1 6 6 5` 59 9 13 7 30 7 20* Significant d'n*rtnci mong tmotung p < 0 05 (j* t#1 *ir 2 d0 Asthma After Being Hired 2 4 3 3 GO --1 CD CD (--> O CD -- CO KiiPlUATCm'f HEALTH IN MAN MADE V1TAE0US FIEEN WOHKEAS 107 ST 0 0 0 6 0 11 an industry with potential respiratory risk: 11% in Plant 3, 31% in Plant 7. Most of those in Plant 7 indicated pre vious sandblasting exposure, although plant management reported no known abrasive blasting with sand in this rural area. Many workers reporting sand blasting possibly referred to plant clean-up jobs, which included blasting with ferrous shot or water (referred to as sandblasting although sand was never used). If sandblasting experience is ignored, the overall percentage re porting such employment is 10%. ranging from 8% (Plant 5) to 16% (Plant 6). TABLE 3 PREVALENCE (V.) OF SELECTED SYMPTOMS AMONG CURRENT SMOKERS BY PLANT Symptom Plant (Fiber Type) Lower Respi ratory Upper Respi ratory Cftromc Bron chitis Shortness Dyspnea at Breath (Grade with >2) __ Wheezing Wheez ing Number of Persons 1 (Ordinary) 81 49 17 7 4 34 toe lOrdlnery) 43 37 17 7 7 22 121 (Ordlnery/tlne) 83 69 25 13 9 39 32 (Ordinary/tine) 32 8 3 4 28 74 (Vary line) JA 46 10 2 4 25 57 Estimation of Worker Exposure The plant manufacturing very fine fi bers (Plant 5) exhibited the highest concentrations of airborne fibers for both OM-sized and EM-sized fibers. For fibers 1 to 3 pm in diameter, the median concentration was 0.210 fibers/mt in Plant 5; all other medians were less than 0.032 fibers/ml. For fi bers less than I pm in diameter, the me dian concentrations were 0.928 fibers/ ml for Plant 5, 0.203 fibers/ml for Plant 3, and less than 0.040 fibers/ml for all other plants. Total accumulated dust exposures exhibited similar trends: median cumu lative EM-sized fiber exposures were 7.4 fibers/ml/yr for Plant 5, 4.3 fibers/ml/yr for Plant 3, and 0.8 fi bers/ml/yr for all other plants. Symptoms Most respiratory symptoms were sig nificantly related to cigarette smoking (table 2), and among current smokers were significantly related to either years or pack-years of smoking. A high prevalence (36%) of Grade I + dyspnea was unexpected in this ac tively employed population. Pulmo nary function results showed that those without dyspnea had higher values for spirometry and lung volume measure ments than those with Grade I dyspnea only, thus validating the responses to this question. Overall, 4 persons re ported Grade 3 dyspnea; none reported Grade 4. Among current smokers, those with previous hazardous employment had greater prevalences of lower respira tory symptoms (65 versus 54%) and wheezing (40 versus 27%), but this in crease was not observed among ex smokers and never smokers. For current smokers, there were dif ferences across the plants for several (Mineral wool) (Mineral wool) Total * Significant pack yaa'al 83 S3 20 84 10 58- 42* 10 10 */nong ?iaiM& at p < 0 06 (j' tat 16 18 42 19 4 24 52 20 a 7* 30* 436 8 df uaing logitHc r*gr#uion. aftar adjusting for symptom prevalences (table 3). which were significant after adjusting for pack-years and previous employment. For 3 symptoms (lower respiratory, wheezing, shortness of breath with wheezing), the elevated prevalences occurred in the mineral wool plants, primarily in Plant 7. By contrast. Plant 7 exhibited very low rates for upper re spiratory symptoms. There were no dose-response rela tionships for any symptoms with any of the quantitative measures of expo sure, either within the mineral wool plants or within the entire population. Moreover, symptom prevalences among ex-smokers and never smokers demon strated no differences between the plants, nor any relationship with ex posure dose. Radiography Overall film quality was excellent, with only 7 of the 948 chest radiographs judged unreadable by 2 of the 3 read ers; these were excluded from all anal yses. Nine films were judged unread able by 1 of the 3 readers; a total of 64 were classified by median judgment as poor quality but nevertheless readable. Using the majority opinion, few films exhibited pleural abnormalities: 15 had pleural thickening, 16 had oblit eration of the costophrenic angle, I had pleural calcification, and I had pleural plaques. These were not associ ated with previous asbestos exposure, and the low prevalences of pleural changes tends to confirm that no sig nificant asbestos was used in these plants. None of the 941 readable films was judged to show large opacities. For profusion of small opacities, 847 (90%) were classified as category 0/0. Of the 94 films (10%) judged to show small opacities, 63 were categorized as 0/1, 25 as 1 /0, and 6 as I /1. None was higher than I/I. Of the 94 films with profusion > 0/1, 55 were categorized by at least 2 readers as having irregular opacities as the primary shape and IJ as having rounded. For the remaining 24, the 2 readers indicating presence of small opacities disagreed on shape. The probability of having small opacities was significantly related to age and cigarette smoking category both for profusion of categories ^ 0/1 and ^ 1/0 (figure 1). There was a significant relationship between presence of small opacities and fiber type category: 8% (22 of 292) Ag* ( fMrl) Fig i. Percentage with profusion of small opacities >0<i, by age and cigarette smoking categories 10% WCIIL HUOH1A, HAMMAD. OUNOMEYER. SHARON. AND JONES *n |oa X -,*0 X **** 9 < 0* n ^ *ft Fig. 2. Pefcenlsge with prolusion ol small opacities > (VI among current smokers, by age and liber lype category. of the radiographs from the ordinary/ fine fiber category were read as profu sion > 1/0, with only 2% (8 of 478), O'?. (0 of 86), and 1V (1 of 85) from the ordinary, very fine, and mineral wool categories, respectively, showing this profusion level. All 6 of the 1/1 films were from the ordinary/fine fiber category. These differences were not explained by gross defects in film qual ity because exclusion of the 64 films of poor quality left the prevalences essen tially unchanged. For both profusion levels (^0/1, ^ 1/0) these increased rates in the ordinary/fine fiber category were observed in both plants in this category for cur rent smokers (for profusion a 1/0, 12^o in Plant 3, 157* in Plant 4, 2V% in other plants combined) but only in Plant 3 for ex-smokers and never smokers {\2h in Plant 3, 1% in Plant 4, l"7o in the others). These differences were significant (p < 0.01) after adjust ing for age, and among smokers, packyears. See figure 2 for the effect among smokers within various age categories. The preceding analyses were also performed for prevalences of small ir regular opacities. Although prev alences were lower than when shape was ignored, exactly the same patterns across plants were observed. For current smokers, presence of small opacities, profusion a 0/1, was also related to several quantitative ex posure variables; these were, in de creasing order of significance: time above the second OM-sized fiber level, duration of employment, cumulative EM-sized fiber exposure, and time above the lowest fiber level (all p values < 0.025). The results for length of em ployment are illustrated in figure 3. The relationships with these exposure variables were significant after adjust ment for age, pack-years, and ordinary/fine fiber type. Moreover, after adjusting for one or more of these ex posure variables, the ordinary/fine fi combined found that the interaction between current smoking and these several exposure variables was signifi cant but that the main effects of the ex posure variables were not. Thus, the dose-response relationship among cur fc+a (|iwl Fig. 3. Prcifttagt with profusion of smalt opacities >Ol among currant smokers. by agt and length of amploymenl. rent smokers was significantly differ ent from the lack of a relationship found among ex-smokers and never smokers. Pulmonary Function ber type continued to be significant (p < 0.01). Although there was an indication of a possible relationship between profu sion ^ 1/0 and employment duration among current smokers, this was not significant, possibly because of the small number (n = 21) of cases. Among ex-smokers and never smok ers, after age and Plant 3 had been ac counted for, no other variables were found to be significantly related to presence of small opacities (only 33 cases ^ 0/1). Because of the observed relationship between small opacities (^ 0/1) and length of employment among smokers, the 163 smokers with over 20 yr em ployment were analyzed separately. Several measures of exposure were sig nificantly related to presence of small opacities (^ 0/1): cumulative EM-sized fiber exposure, average EM-sized fiber concentration, average total fiber con centration, and time above the second OM-sized Tiber level (table 4). Although all regressions allowed for a possible effect of previous hazardous employment (never close to statistical significance), all analyses were re peated excluding persons with such em ployment; no changes in the overall re sults were observed. Analyses of the smoking groups Of the 897 employees who participated in lung function testing, 836 had satis factory spirometry, 824 repeatable lung volume measurements, and 796 repeatable pulmonary diffusing capaci ty results. Percentages of predicted (based on nonsmoker values) were calculated for each person using published equations (9-12) that take into account age, height, and race. Mean percentages of predicted (table 5) were generally high, with significant differences between the cigarette smoking groups. Among current smokers, there was a signifi cant correlation with pack-years of cigarette smoking. For all symptoms, those with the symptom had significantly (p < 0.001) lower spirometric measurements and higher lung volumes than those with out the symptom, after adjusting for age, height, race, and smoking. To assess the relationship between lung function and profusion of small opacities, residuals were obtained from internal regression equations (using age, height, race, and 2 smoking cate gory-by-age interaction terms); FEV,, FEFji-tj, FEV./FVCft, and Dlc0 exhibited a significant trend in mean residuals by profusion of small opaci ties (table 6). After adjusting for pro fusion, there were no significant diffef*'"5 ences in pulmonary function by prF"* CD CD ^ TABLE 4 PERCENTAGE WITH SMALL OPACITIES > (VI BY AGE, FIBER TYPE CATEGORY, AND CUMULATIVE ELECTRON MICROSCOPY (EM) EXPOSURE; CURRENT SMOKERS WITH > 20 YEARS EXPOSURE CD CD Ag (yr) < 0.1 Ordlnery/FIne Plants Exposuit Lavsi* 0.1 -08 > 0 ft < 0.1 Other Plants Exposure Level* 01-08 >08 38-45 48-55 > 55 Alt ages 015) - (n = 0> 13 (*> 25 (2/8) 29 (2/7) 22 (5/23) 0 10/1) 43 (10/21) T5 (9/12) 56 (19/34) 9 uni) 6 (2/24) 14 (1/7) 10 (4/42) 0 (03) 15 (3/20) IS (2/11) 15 (5/34) 0 10/4) 16 (2/11) 40 (5/15) 27 (8/30) MWIMTOftV HEALTH IN MANMADE VTTXEOUS FIAEA WOAKEM lot ' STOODG0J9 TABLE 5 PULMONARY FUNCTION (MEAN PERCENT OF PREDICTED) BY CIGARETTE SMOKING CATEGORY CJgaralta Smoking Category Function n Navef Ex Currant FVC FEV, FEV,/FVC. V.f FEFjm RV VC TLC Ol DUALVOL 836 no* 110 836 109* 106 836 80' 77 830 94' 83 824 114* 130 824 112" 111 824 108* 111 798 121' 118 796 117* 111 106 98 75 75 137 107 111 104 99 OtHnltton ot *tbf+v'*uonf FVC a fo*C*d H*i ep*C tf, PEV, forced ap4riory *otum# in 3 forevd tipiftiory flow duf'Ag m miOdie half of (ho FVC. RV = fttiduai voiomo. VC s vital capacity; TLC 3 total lung capacity. IX 3 diffusing capacity, DUALVOL = ratio of dftfuamg capacity to ahraoiar velum*. * Significant diffa/anca among smoking c*l*go<>* by analyala of vananca, p < OOOt t Mtan valua in aach smoking caiagory ^ Significant dlflaranct among smoking catagoiat by anaiyata of vananca, p < 0 02 mary type of small opacities (rounded versus irregular). For current smokers, comparisons across plants revealed that after adjust ing for known factors (age, height, race, pack-years), there were significant dif ferences for some of the lung volumes (RV, TLC) and for diffusing capacity (Dl, Dl/ALVOL), but that these dif ferences showed no consistent or meaningful pattern. For example, Dl and Dl/ALVOL were lower in Plant 3 than in the other plants, but none of the other functions were, whereas RV was high in Plants 5 and 7, but all other functions were comparable. There were no differences for the spirometric measurements, and similar analyses of ex-smokers and never smokers found no differences for any functions. These results do not support the hypothesis that fiber type or plant variables are having important effects on pulmonary function among current smokers. Combining the smoking groups, FEVt, FEF-m, and FEV,/FVC% were found to be significantly related to length of employment and time above the lowest fiber concentration level, after adjusting for known fac tors, including smoking effects. The coefficients of these variables, how ever, were positive, indicating increas ing function with increasing exposure time (employment time coefficients: 0.009 L/yr for FEV,, 0.019 L/s/yr Tor FEFjj-jj, 0.140^o/yr for FEV^ FVC^o). Detailed analyses of residuals and weighted regressions failed to ac count for these findings. When these regressions were performed on the smoking categories separately, no sig nificant relationships (p values > 0.10) were found. Analyses for RV, RVCTLC, and TLC showed similar results: coefficients were negative, indicating a decrease in these functions with increasing expos ure time; these results were generally consistent for the smoking groups. No exposure relationships were found for the measures of diffusing capacity. Discussion Assessment of the respiratory health of 1,028 workers engaged in the produc tion of MMVF gave the impression of a generally healthy population experi encing no substantial lung disease. In all studies, large unexplained vari ability in symptom prevalences, as determined by standardized question naire, occurred both within and be tween surveyed populations. For this table s MEAN RESIDUALS' FOR SELECTED FUNCTIONS BV PROFUSION OF SMALL OPACITIES Profusion of Small Opacities FVC (L) era (M9)t (Vi (51) 1/0 (20) 1/1 (5) 0.003 (0.566)* -0.059 (0 891) 0.014 (0 5961 -0 878 10 8431 FEV, (D 0.010 (0.581) -0.082 (0 573) -0.115 (0 534) -0 653 (0.554)4 FEF,,.,, (L/t) 0.025 (1.081) -0 153 (0.797) -0 274 (0 993) -0 783 (1 002)4 FEV^FVC <%) 0.117 (7 203) -0 494 (7 805) - 2 737 (10.672) -3940 (8 187)4 RV H) 0.004 (0.507) -0.045 (0 525) 0.091 (0.767) 0.079 (0.450) TLC (O 0.014 (0 782) -0.090 (0.999) 0.117 (1 069) -0.625 (1.091) Dlco (mUmlru mmHg) 0.093 (4.859) -0.905 (4 929) - 1 881 (8 517) - 1.433 (8 405) ' Otsarvud grtCictatJ (using gt. haignt. >aca, and 2 smoaingDs aga Inlarachon varlatMaal * Number in category *>tn good sonomalry * Standard daviition * Significant gradient in mean residuals sc rasa profusion cal agones ty simple linear regression on categories coOed '2.J.4. p < 0 024 reason it is impossible to firmly estab lish the "background" prevalences that can be expected in a working pop ulation not occupationally exposed to any respiratory risk. Despite this diffi culty, it is unlikely that any of the ob served symptom prevalences were sig nificantly elevated in this population. Indeed, the prevalence of bronchitis seemed low, particularly among never smokers and ex-smokers. The relation ship of symptom prevalences with pul monary function measurements and cigarette smoking, as well as with the level of smoking among current smok ers. supports the validity of the ques tionnaire responses. Although an increased prevalence of certain symptoms was detected among current smokers in the 2 mineral wool plants, statistical significance was at tained in only 1 of these plants, and no dose-response relationships were found. Bronchitis rates were not elevated in these plants and the nonsmokers exhibited no increased symptoms. It is concluded that prevalences of respiratory symptoms are not increased in this population and no coherent pattern of symptoms in relation to ex posure to MMVF emerges. Lung function testing also failed to detect any adverse effect of exposure, although measurements were related to age, height, race, cigarette smoking status, and level of smoking. Using published equations to provide exter nal comparisons, mean values for all smoking groups were near or above lOO^o of predicted. After taking into account other in fluencing factors, there was no adverse association between functional meas urements and any of the indexes of exposure to MMVF. In fact, after age and smoking were taken into account, (he associations of RV and expiratory flow rates with length of exposure suggested a possible "healthy worker" effect. The negative findings concerning symptoms and pulmonary function testing in this study are consistent with the generally reassuring results of all reports published to date. Several sur veys of working populations failed to detect any chronic respiratory effects of exposure to MMVF (13-16). The single histologic study also reported negative results, with lungs of exposed workers exhibiting no differences from those of matched control subjects (17). Although some mortality studies showed 110 WEILL, HUOHEJ, HAMUAD. QLIMOMEYEA. SMAAON. ANQ JOMCA an increased risk of nonmalignant respiratory disease, there was no in dication of increased lung cancer risk (3, 4, 18-20), and animal inhalation studies produced adverse respiratory effects only after either intratracheal injection or direct installation of fibers into the pleural space (21-25). However, reported findings of new and updated studies presented at the WHO International Occupational Health Conference on the Biological Effects of Man-Made Mineral Fibres (Copenhagen, 1982), while providing no cause for alarm concerning health risks from MMVF, were not entirely negative. In a U.S. mortality study, there was no overall excess of lung cancer among 14,884 fibrous glass workers, but an increase was observed (47 observed, 36.0 expected) among those with over 30 yr of follow-up from initial expo sure (5). The 1,846 mineral wool workers in this study exhibited an over all excess of lung cancer mortality (45 versus 28 expected). Both groups of workers were found to have an increased risk of death caused by nonmalignant respiratory disease. A European study of 17,083 employees, although not completed, reported a significant ex cess of lung cancer deaths (12 versus 5.6) among those followed over 30 yr (26). Similar results were reported from another U.S. study (4,394 work ers), which found an excess of lung cancer deaths (11 versus 6.2 expected) among workers with over 20 yr of employment and 30 yr of follow-up (27). No mesotheliomas were found in the U.S. studies, but one was found in the European studies. It is difficult to interpret the results of these mortality studies with regard to exposure to MMVF because none found a dose-response relationship and none had smoking information avail able. The past use of asbestos in some of the U.S. mineral wool plants intro duces further uncertainty regarding these results. Recent animal experiments with MMVF confirmed the results of earlier studies: MMVF injected intrapleurally or intraperitoneally produced mesothe liomas, but none was produced by in halation. Inhalation studies with glass fibers produced no or mild fibrosis; fewer lung tumors occurred in exposed rats than in unexposed control animals (28-31). Inhalation of very high con centrations of ceramic fibers produced interstitial fibrosis and neoplasms in both plants. They have produced fine the lung tissue of animals (29). Several fibers for a sufficiently long period to studies showed that pitting, corrosion, make the detection of an effect pos progressive shell formation, decreasing sible. Among ex-smokers and never fiber diameter, fracturing, and dissolu smokers, this effect was observed only tion of MMVF occur in vivo and in in the dustier plant. Thus, in the gener vitro (28. 30, 32-34). These findings ally less dusty plant, an increased pre concerning the durability of MMVF in valence occurred only among smokers, lung tissue may help to explain the dis whereas in the dustier plant, an in crepancies reported in the literature creased prevalence was observed in all between intrapleural injection and smoking categories. If the higher dust inhalation studies. levels contributed to the increased A recent cross-sectional survey of prevalence of small opacities, then a 340 employees of a U.K. fibrous glass fiber dose-response relationship should plant reported that 13V* of the chest be demonstrable in this fiber type cate radiographs of men exhibited small gory after smoking has been taken into opacities of category 1/0 or higher account. In fact, such a relationship (35). Although this study had consider was found only among current smokers. able technical limitations, the presence This could be a result of the smaller of small opacities was related to previ number of ex-smokers and never ous employment in dusty trades and smokers with small opacities, but it general environmental exposures, but may also reflect the recognized diffi there was no relationship with expos culties of individual exposure recon ure to MMVF as measured either by struction. length or intensity of exposure. Al If the increased probability of small though FEV, and VC were low, neither opacities is attributable to exposure to functional measurements nor symp fibers with the smallest diameter, then toms were related to exposure to the plant manufacturing very fine MMVF. Two smaller studies of mineral fibers, whose average fiber concentra wool workers found no evidence of tion levels were considerably higher adverse functional effects (36, 37). than all other plants, would be ex-, The present study, also presented at pected to exhibit increased prevalence^ the Copenhagen Conference, while of small opacities. In fact, they were finding no adverse exposure effects as very low in this plant. Because thestf' measured by symptom prevalence and function testing, found that the pres ence of small opacities was related to workers were the youngest (mediaTP age, 34 yr) and overall had the shortcP employment duration (a median of T3 some measures of exposure. yr; only 14 employed for more than IT" Classification of chest radiographs yr), longer follow-up is required IT demonstrated a low prevalence of fully assess radiographic outcome ijT small opacities, with no large opacities. this exposure category. Films considered to show small opaci The focus of environmental charac ties were almost exclusively in the terization, and therefore exposure lowest profusion categories (0/1 or reconstruction, has been on qualitative 1/0) on median reading. On individual and quantitative aspects of airborne readings, a choice of these ILO cate fibers in this study. Although this ap gories indicates that the reader favored proach seems well based in prioritizing (0/1) or seriously considered (1/0) a concern about potential respiratory negative judgment. health effects, we recognize that Prevalence of small opacities rose as further exploration of nonfibrous par both age and smoking increased, find ticulates and gaseous inhalants in this ings that have been reported elsewhere industry is a logical next step. (38-40). Among current smokers, , Inevitably, attention will be directed probability of small opacities was posi-' to the potential biologic significance of lively associated with exposure dura the correlations found in this study tion; however, among ex-smokers and between low profusion of small opaci never smokers, there was no associa ties and some indexes of exposures. tion with exposure measures. Possibly there is minimal dust accumu With respect to fiber type, exposure lation, retention, and persistence of to a mixture of ordinary and fine fibers MMVF with fine diameters in lung tis significantly increased the probability sue. If so, these effects, in general, are of low-level small opacities among cur attaining the radiographic detection rent smokers, an effect observed in limit only in current smokers, perhaps MWIMTOirr HEALTH IN MAN-MADE VTTACOU3 FIDEH WOHKEHS lit ST 00 0 60 21 because of less effective clearance of particulates. The absence of intermedi ate or higher profusion of small opac ities, in a population with a consid erable range of exposure durations, suggests that a progressive, diffuse 6. Esmen NA, Hammad YY. Log-normality of environmental sampling data (abstract). J Envi ron Sci Health 1977; 12:29-41. 7. Ferris BG. Epidemiology standardization project. 11. Recommended respiratory disease questionnaires for use with adults and children in epidemiologic research. Am Rev Respir Dis 24. Poll F, Huth F. Friedrichs KH. Results of animat carcinogenesis studies after application of fibrous glass and their implications regarding human exposure. In: Occupational exposure to fibrous glass. Proceedings of a Symposium. Col lege Park: National Institute for Occupational Safety and Health, 1974:183-92. tissue reaction (e.g., fibrosis) is un 1978; 118(Suppl):7-33. likely, yet it cannot be excluded. Because interobserver variability is greatest at these lowest profusion levels 8. Guidelines for the use of ILO international classification of radiographs of pneumoconioses. Rev. ed. Geneva. International Labor Office, 1980. Occupational Safety and Health Series, no. (0/1, 1/0) and because of the im 22, (rev. 80). portant influences of age and smoking 9. Boren HG, Kory RC. Syner JC. The -on the probability of detecting thesis Veterans Administration-Army cooperative opacities, the presence of small opac ities at these levels has a low order of specificity for an occupational expo study of pulmonary function. Am J Med 1966; >11:96-114. 10. Arcangeli P, Cotes JE, Cournind A, edi. /formal values for respiratory function in man. sure effect. Clearly, the findings re 'Alghero. Italy: Panminerva Medica, 1969 335 ported here should be considered pre 11. Arcangeli P, Cotes JE. Cournand A, eds. liminary, requiring replication and Normal values for respiratory function in man. confirmation before firm conclusions Alghero, Italy: Panminerva Medica. 1969 336. can be drawn. The suggestion that fibers with fine diameters may be hav ing an effect on lung parenchyma makes it prudent to continue biologic surveillance of workers exposed to 12. Rossiter CE. Weill H. Ethnic differences in lung function, evidence for proportional differ ences. Int J Epidemiol 1974. 3:55-61. 13. Wright GW. Airborne fibrous glass par ticles: chest roentgenograms of persons wiih pro longed exposure. Arch Environ Health 1968; MMVF with small and very small 21:175-81. 25. Wright GW, Kuschner M. The influence of varying tcnghls of glass and asbestos fibers on tissue response in guinea pigs. In: Walton WH, ed, Inhaled panicles. Proceedings of an inter national symposium organized by the British Occupational Hygiene Society, Edinburgh, September 22-26, 1975. Oxford: Pergamon Press, 1977:455-72. 26. Saracci R, Simonato L, Esteve J. er at. Mor tality and cancer incidence of man-made mineral (vitreous) fibre production workers in seven European countries. In: Biological effects of man-made mineral fibres. Proceedings of the Oc cupational Health Conference 1982. Copenha gen: (in press). 27. Morgan RW. Kaplan SD, Bratsberg JA, Mortality in fibrous glass production workers. In: Biological effects of man-made mineral fibres. Proceedings of the Occupational Health Conference 1982. Copenhagen: (in press). 28. McConnell EE, Wagner JC, Skidmore JW, Moore JA. Comparable effects of inhalation USA/UK. In: Biological effects of man-made mineral fibres. Proceedings of the Occupational diameters. This approach is particu larly indicated when a working popula tion may have had relatively short-term exposure on average and the potential anticipated biologic response is sus 14. Utidjian HMW, deTreville RTP Fibrous glass manufacturing and health report of an epidemiological study: Parts I and II In. Pro ceedings of the 35th Annual Meeting of (he In dustrial Health Foundation. 1970. Pitlsburgh: Indusiria! Health Foundation, 1970 Health Conference 1982. Copenhagen: (in press). 29 Davis JMG. Bolton RE. Donaldsoo K, Jones A, Wright A. The effects of inhalation of ceramic aluminum silicate fibres in rats. In: Bio logical effects of man-made mineral fibres. Pro pected of having a long latency period. 15. Nasr ANM, Ditchck T. Schollens PA The ceedings of the Occupational Health Conference In general, however, the minimal evi dence of respiratory effects detected in this investigation, which cannot, at present, be considered clinically signifi cant, is encouraging concerning the presence of radiographic abnormalities in the chests of fibreglass workers. J Occup Med 1971, 13:371-6. 16. Hill JW, Whitehead WS. Cameron JD, Hedgecock GA. Glasi fibres absence of pul monary hazard in production workers Br J Ind 1982. Copenhagen, (in press). 30. Pott F. Schlipkoter HW, Ziem V. Recent results of implantation studies. In- Biological effects of man-made mineral fibres. Proceed ings of the Occupational Health Conference 1982. Copenhagen- (in press). question of potential health effects of Med 1973; 30:174-9. 31. Let KP, Reinhardt CF. Biological studies exposure to MMVF. 17. Gross P, Harley RA. David JMG. The on other MMMF In. Biological effects of man lungs of fiber glass workers: comparison with the made mineral fibres. Proceedings of the Occupa lungs of a control population In: Symposium on tional Health Conference 1982. Copenhagen- (in Ralerencas Occupational Exposure to Fibrous Glasi. Col press). 1. U.S. Department of Health, Education and Welfare. Criteria for a recommended standard: occupational exposure to fibrous |lass. Washing ton D C.: U.S. Government Priming Office. 1977. (DHEW Publication No. [NIOSH] 77- 152.) lege Park: National Institute for Occupational Safety and Health, 1974:147-57. 18. Bayliss DL, Dement JM, Wagoner JK, Blejer HP Mortality patterns among fibrous glass production workers. Ann NY Acad Sci 1976; 271324-35. 32. Hammad YY. Deposition and elimination of MMMF In: Biological effects of man-made mineral fibres. Proceedings of the Occupational Health Conference 1982. Copenhagen: (in press). 33. Klingholz R. Steinkopf B. The reactions of 2. Esmen N, Corn M, Hammad Y, Whittier D, Kotslti N. Summary of measurements of em ployee exposure to airborne dust and fiber in uateen facilities producing man-made mineral 19. Morgan RW. Kaplan SD Mortality study of fibrous glass production workers. Arch En viron Health 1981; 36:179-83. 20. Robinson CF, Dement JM, Ness GO, Wax- MMMF m a physiological model fluid and in water. In: Biological effects of man-made mineral fibres. Proceedings of the Occupational Health Conference 1982. Copenhagen: (in fibers. Am Ind Hyg Assoc J 1979; 40:108-17. weiler RJ. Modality patterns of rock and slag press). 3. Enterline PE, Henderson V. The health of retired fibrous glass workers. Arch Environ Health 1975; 30:113-16. mineral wool production workers: an epidemio logical and environmental study Br J Ind Med 1982 ; 39:45-5). 34. Leinewtber JP. Solubility of fibres in vitro and in vivo. In: Biological effects of man-made mineral fibres. Proceedings of the Occupational Enterline PE, Marsh GM. Mortality of 21. Gross P. deTreville RTP. Crailey LJ. Health Conference 1982 Copenhagen: (in Orkers in man-made mineral fiber industry. In: Granquist WT, Pundsack FL. The pulmonary press). **ner JC. ed, Biological effects of mineral response to fibrous dusts of diverse composi 35. Hill JW. Rossiter CE. Foden DW. A pilot fibres, Vol. 2. Lyon: International Agency for tions. Am Ind Hyg Assoc J 1970; 31:125-32 respiratory morbidity study of workers in an Research on Cancer. Scientific Publications no. 22. Gross P. Kaschok M, Tolkes EB, Babyak MMMF plant in the United Kingdom. In Bio 30. 1980:965-72. MA, deTreville RTP. The pulmonary reaction to logical effecu of man-made mineral fibres Pro 3- Enterline PE. Marsh GM The health of high concentraiions of fibrous glass dusl. Arch ceedings of the Occupational Health Conference orkers in the U.S. mineral fiber industry In: Environ Health 1970; 20:696-704 1982. Copenhagen: (in pressl. Biological effects of man-made mineral fibres. 23. Stanton MD. Wrench C. Mechanisms of 36 Malmberg P. Hedenstrom H, Kolmoden- Proceedings of the Occuparional Health Con- mesothelioma induction with asbestos and fi Hedman B. Krantz S. Pulmonary function in f*tence 1982, Copenhagen (in press) brous glass. J Natl Cancer Inst 1972, 48 797. workers in a mineral rock fiber plant In: Bio- 112 WUU. HU0HM. HAMMAO. OUNOMCTfR. (MAHON. ANO JOMCg logical effects of min-made mineral fibres. Pro ceeding of the Occupational Health Conference 1982. Copenhagen: (in press). 37. Skuric Z. Stahuljak-Beritic O. Occupa tional exposure and ventilatory function changes in rock wool workers. In: Biological effects of man-made mineral fibres. Proceedings of the Occupational Health Conference 1982. Copen hagen: (in press). 38. Weiss W. Cigarette smoking, asbestos and pulmonary fibrosis. Am Rev Respir Dis 1971; 104:223-7. 39. Amandus HE. Lapp NL. Jacobson G. Reger RB. Significance of irregular small opac ities in radiographs of coal miners in the USA. Br J Ind Med 1976; 33:13-17. 40. Rossittr CE, Harries PG. UK naval dock yards asbestos study: survey of the sample popu lation aged 30-39 years. Br J Ind Med 1979; 36: 281-91. ST0006022 B073 RM l 30 P FARO L*b INVfc ST 37* IJ02-1-6K*"/Hi/4904*inTSir. UO'U LaBORATOKV IN V ESTIMATION Popvruhl ` 19K.I h\ rhe l'ruled Stale* ('anWi*n I)tw K>n of the Jntrrnationa) Atjulerm of // Editorial Vol 49. Sn 4. p IT9. I9K.I f'rtntnj in / * n .4 How Do Mineral Dusts Cause Lung Injury? The chronic inhalation of silica or asbestos particles can lead to disabling pulmonary fibrosis. In addition, asbestos exposure is associated with at least two types of malignant tumors. By contrast, chronic inhalation of native carbon particles, such as soft coal or the soot from the burning of fossil fuels, is accompanied neither bysignificant pulmonary fibrosis nor bv increased risk of neoplasia. It is now generally accepted that this difference between the response to fibrogenic minerals. e.g., silica and asbestos, and relatively inert carbon particles is a consequence of the ability of the former to kill the cells that engulf them. Analysis of the mechanisms of silica and asbestos cytotoxicity has. therefore. become central to an understanding of the pneumoconioses that result from exposure to these and other mineral dusts. Silica and asbestos present unique problems to the experimental toxicologist. The chemical basis for their biologic activity is not readily apparent. In addition, unlike other toxins that are distributed homogeneously in biologic media, silica and asbestos are insoluble minerals, the physical properties of which affect their biologic activity. Cytotoxicity seems to result from contact between silica and asbestos particles and target cells. For many years, the killing of these cells has been related to the fate of the particles after this contact. Alveolar macrophages are the major cells that phagocytose the mineral dusts upon their inhalation. The release of lysosomal enzymes into the cytosol follows rapidly U. (1, 7, 9l. The breakdown of cellular components as a result of the release of lysosomal components has been hypothesized to produce irreversible cell injury In). Although intracellular lysosomal rupture has been documented in silica- and asbestos-treated macrophages, the evidence that it causes cell death is only circumstantial. Two conditions must be met before a causal relationship between lysosomal rupture and cell death can be inferred. Lysosomal release must occur before cell injury becomes irreversible, and the mechanisms whereby such release leads to irreversible injury must he identified. In the case of most hepatotoxins, rupture of Ivsosomes occurs after irreversible injury has developed 110). In these situations, the late release of lysosomal enzymes simply accompanies lysis or enzymatic removal of dead cells. Bycontrast, lysosomal rupture is temporarily related to cell death in the toxicity of silica and asbestos to macrophages. However, in these cases the biochemical and functional consequences of intracellular lysosomal rupture have not been identified. Furthermore. Kane et al. (8) were able to dissociate intracellular lysosomal rupture from the subsequent cell death after exposure of macrophages to silica particles. Without extracellular calcium ions, phagocytosis of silica particles occurred, followed by intracellular lysosomal rupture. However, all of the cells remained viable over the time course that cells exposed to silica particles, in the presence of extracellular calcium ions, were killed. In addition, lvsosomal rupture was not associated with measurable degradation of total DNA, RNA, protein, or phospholipid, even in the presence of extracellular calcium ions. This finding suggests that intracellular lysosomal rupture is not related to the cell death caused bv silica and, presumably, therefore, also bv asbestos. With silica, cell death was dependent on extracellular calcium ions and seemed to be a consequence of an influx of these ions across the permeability barrier at the*-^ plasma membrane, which was directly damaged by**"l exposure to these particles. *--J Such a conclusion places the problem of cell injury*--* induced by mineral particles in the same context as the*-- study of the mechanisms of action of other hazards. It is^ becoming increasingly clear that disturbances in plasma*--13 * membrane function are responsible for the genesis ofr' irreversible cell injury with most toxins (41. The study^0 of mineral particle-induced cell injury should, therefore, increasingly focus on the effects such agents have on cellular membranes. In this issue of Laboratory Investigation. Brody, Berwyn, and Hill 12) extend our understanding of the mechanism of the interaction of asbestos fibers with the erythrocyte membrane. Red blood cells lyse after exposure to asbestos or silica, and the release of hemoglobin can be readily quantitated as an index of membrane damage. Red cel! hemolysis is, therefore, a convenient experimental model with which to study the effect of mineral particles on cellular membranes. Brody et al. (2) have studied the interactions of chrysotile and crocidolite asbestos with red blood cell membranes. Data are provided to support the hypothesis that the membrane disruption by asbestos is secondary to an initial interaction between positive charges on the fiber surface and negatively charged sialic acid residues on the membrane. Previously, it was shown that treatment of erythrocytes with neuraminidase, an enzyme that removes sialic acid from the cell surface, reduces the hemolytic activity of chrysotile asbestos (6). Chrysotile. but not the less active crocidolite asbestos, has now been shown to prevent the neuraminidase-mediated removal of sialic acid from red blood cells, implying that the I 1 GHT U-* lLuU^ o iUKlNj LJuAL I 1 AUHEL 380 FARBER Laboratory Investigation interaction between the chrysotiie and the membranebound sialic acid groups prevented the enzyme from removing them. The authors then used a second assay to obtain a similar result. Wheat germ agglutinin is known to bind to /V-acetyl neuraminic (sialic) acid and JV-acetyl glucosamine. Wheat germ agglutinin bound to red blood cell membranes can be visualized, if it is first conjugated with colloidal gold particles. The number of gold spheres bound to a membrane can then be determined with scanning electron microscopy. Brody et al. (2) show that the pretreatment of red blood cells with chrysotiie, but not with crocidolite, asbestos reduced the number of Au-wheat germ agglutinin-labeled sites to less than 30% of the control level. On the basis of the present and previous studies, it may be presumed that the interaction between asbestos fibers and negatively charged sialic acid residues on the red cell surface is followed by an increase in the permeability to sodium and potassium ions. An influx of sodium and accompanying water molecules would then lead to hemolysis. How does the interaction with asbestos fibers change the permeability of the red cell membrane'' Brody et a!. (2) do not provide any data relative to this critical question but do suggest a possible mechanism, namely, redistribution of membrane glycoproteins. Aggregations of these proteins could create new ion channels. Other possible mechanisms include an increased ordering of the phospholipids adjacent to the bound asbestos fibers. Altered lipid-protein or lipid-lipid interactions could affect the permeability of the membrane. Additional studies are needed to distinguish between these potential mechanisms. It also remains to be shown that the findings regarding the nature of the interaction of asbestos fibers with erythrocytes are relevant to the effects on macrophages. The findings of Brody et al. (2) are significant in that they serve to focus future investigations sharply on critical questions. Answers to such questions should bring us closer to a definitive understanding of the mechanisms of cell injury with this important and interesting group of environmental contaminants. John I,. Farber Department of Pathology and Laboratory Medicine Hahnemann University School of Medicine Philadelphia, Pennsylvania REFERENCES 1. Allison AC. Harmgion )S. Birbeck M: An examination of the cytotoxic effect* of silica on macrophages. J Exp Med 124:141.1966 2. Brody AH. Herwvn C. Hill l.H: Interactions of chrysotiie and crocidolite asbestos with red blood cell membranes. 1-ab Invest 49:46*. 19*1 .1 Davies P. Allison AC. Ackerman -I. Butterfield A. Williams S' Asbestos induces selective release of lysosomal enzymes from mononuclear phagocytes Nature J.S 1:42'1* 1974 4. Earlier JL Memhrane injury and calcium homeostasis in the path ogenesis of eoagulative necrosis l-ab Invest 47 114. 1982 .r> (loldstein IM. Wei**man C Intracellular digestion: lysosomes and cellular in/urv In Handbook of Physiology, edited by Lee. I)HK. Vol 9, p 64LI Bethesda. Maryland. American Physiological Societv. ty77 6. Harmgton -IS. Miller K, Mac nab (. Hemolysis hy asbestos. Envi ron Res 4 y.S, 1971 7 Jaurand M(\ Magne L. Boulmer JL. Bignon *J In vitro reactivity of alveolar macrophages and red blood cells with asbestos libers treated with oxalic acid, sulfur dioxide and lienzo- 1,4-pvrene. Tox icolugy 21 M'A. 19H1 H. Kane AB. Stanton RP, Raymond E(. Dohson ME. Knalelc ME. Farber JI. Dissociation of intracellular lysosomal rupture from the cell death caused hv silica. J (`ell Biol H7.64M. 19*4) y. Nadlrr S. OoldfWher S The intracellular release of lysosome contents tn macrophage* that have ingested silica. -I Histoehem Cytochem 1*. 16*. 1970 10. Slater TF, Dreenhaum AL Thanges m lysosomal enzymes in acute experimental liver mjurv Biochem I 96 464. I96.S ST 0 0 ) S0 2