Document ppryowRvOjknnoNw1DXnm0KRX

- V 253 Low Exposure to Asbestos Gas Exchange in Ship Pipe Coverers and Controls Raymond L. H. Murphy, Jr., MPH, ScD, MU, and Edward A. Gaensler, MD, Boston; Ralph A. Redding, MD, Providence, Rl; Roger Belleau, MD, Quebec; Patrick J. Keelan, MD, Dublin; and Arthur A. Smith, MSEE; Anne M. Goff, MD; and Benjamin G. Ferris, Jr., MD, Boston In a previous survey of shipyard pipe coverers and controls, clinical criteria were used to define pulmonary asbestosis. This second survey focused on physiologic abnormalities. Pipe coverers had significantly reduced vital capacities (FYC) as well as single breath (DlSB) and exercise steady state (DlSS-Ex) diffusing ca pacities. Resting steady state diffusing capacity (DlSS-R), fraction carbon monoxide removed (Fco), and the diffusing constant (K) differed from selected normals but not from all controls. Airways resistance, specific conductance, minute ventilation, arterial carbon dioxide pressure (Pco3) and dead space were not significantly abnormal. Obstructive disease was equally com mon in both groups. Alt workers with clinical "asbestosis" also had severely reduced Dl and Fco. The significance of isolated reduction of DL requires further study. However, in a third survey three years later, Dl in exposed workers had de teriorated more rapidly than FVC; some with initially isolated reduction of Dl had developed other signs of disease. tosis based on clinical criteria.1 We were less certain that a "negative" diagnosis indicated absence of interstitial lung disease. It has been suggested that tests of respiratory gas exchange may be more sensitive in detecting early, or subclinical asbestosis,2 0 but their importance remains controversial.7 3 Our defined population presented an op portunity for evaluation of such tests. Expo sure had varied from 1 to 30 years, and therefore we expected all gradations from slight to advanced disease. While in the past, quantitative dust exposure histories have been difficult to obtain, exposure in our pipe coverers had been monitored.1 Finally, a control group, comparable except for dust exposure, has not been available previously. Accordingly, this second survey was planned to explore the usefulness of tests of respira tory gas exchange in assessing the spectrum of pulmonary asbestosis. Population Studies In a shipyard survey in 1965, we felt confident of a "positive" diagnosis of asbes- Submitted tor publication Dec 17, 1971; accepted March 13.1972. From the Thoracic Services (Drs. Redding, Bel leau, and Keelan, and Mr. Smith), and departments of medicine (Dr. GoS) and surgery (Dr. Gaensler), Boston University School of Medicine, and the Department of Physiology, Harvard School of Pub lic Health (Drs. Ferris and Murphy), Boston. Dr. Redding is now with Brown University Medical School; Dr. Belleau, Laval University; and Dr. Keelan, National University in Dublin. Reprint requests to Boston University School of Medicine, 80 E Concord St, Boston 02118 (Dr, Gaensler). In the first survey, we studied all 101 pipe coverers employed in a New England ship yard. An equal number of controls, matched for duration of employment and age, were selected from among shipfitters and pipe fitters without known exposure to asbestos. Participation among the latter was volun tary and seven refused, leaving 94 controls. In this survey, one year later, three of the pipe coverers with "asbestosis" had died, and 14 others had left for lack of work: the latter were younger men, none of whom had been diagnosed as asbestotics. Among the Arch Environ Health--Vol 25, Oct 1972 j[JJB 00 1 5833 *7 254 rLOW EXPOSURE TO ASBESTOS--MUHmY ET AL controls, lack of work, vacations, retire* ments, and three refusals accounted for 24 losses so that there remained 84 pipe coverers and 70 controls for this second survey. Smoking habits were similar: 66.4% of the pipe coverers and 68.1% of the controls were present smokers.1 Dust exposure, under surveillance for 20 years, had been at or near the then recommended threshold limit value of 5 million particles per cubic foot (mppcf).1 In the first survey, asbestosis was defined by tire presence of three or more of five clinical abnormalities including (1) moder ate dyspnea (on climbing one flight of stairs or less, approximately equivalent to Fletcher grade III); (2) crepitant basilar rales in two or more sites; (3) clubbing of the fingers (a hyponychial angle of 198 or greater); (4) a vital capacity of less than 80% of predicted; and (5) a roentgenogram consistent with moderately advanced (code 5) or advanced asbestosis (code 6). These codes 5 and 6 correspond to the new UICC/Cincinnati classification of profusion 2 and 3.* By this definition, asbestosis was diagnosed in 11 pipe coverers and one control; it was found first after 13 years of exposure or 60 mppcfyears; with more than 20 years of exposure, the prevalence was 38%.1 Among the three of 11 pipe coverers with clinical asbestosis who were not seen in the second survey, one had died of coronary occlusion and postmortem examination showed asbestosis; autopsy in another showed asbestosis and cor pulmonale; and the third died of bronchopneumonia without autopsy. Four pipe coverers had to be classi fied as new cases of asbestosis on the basis of our clinical criteria. Methods The vital capacity (FVC) and the volume exhaled during the first second (FEVj) were measured in both surveys.1 Methods of plethys mography and measurement of diffusing capac ity have been described1*1-1# and were modified as follows. Body Plethysmography.--The functional re sidual capacity (FRCbp), airways resistance (Ra), and specific conductance (Ga/Lv) were measured with a constant volume, variable pressure box.1# Calibration and loops during panting were displayed on a storage oscilloscope, and the angles of the slopes were read directly. Steady State Observations.--The breathing circuit consisted of an Otis-McKerrow valve with a baffle to reduce dead space to 40 mL Sub jects inspired humidified 0.1% CO in air and expired into a balloon within a 200 liter steel drum with a window. This drum was connected to a spirometer with a kymograph to monitor minute ventilation (Ve) and respiratory frequen cy (f). End-tidal C02 partial pressure (pBtco.) was monitored at the mouth. Mixed expired gas was analyzed after each run while a small blower emptied the balloon within the drum. At rest, subjects were seated on a treadmill, adequate relaxation, and a steady state were indicated by steady (PETea.) and f. Then, the CO mixture was breathed for six minutes, with expired gas collection for the last three'min utes. For exercise, the breathing assembly was raised and the workers walked at 2 mph at an 8% incline for six minutes, breathing the CO mixture for the last four minutes, and with gas collection for the last two minutes. This moder ate exercise increased Oo uptake about fourfold. Calculations--Minute ventilation (Ve) and tidal volume (Vt) corrected to BTPS, 02 up take (Vo,,), C02 output (Vco2), and CO uptake (Vco) all corrected to STPD. The ventilation equivalent for On (VeO.,) and the respiratory exchange ratio (R) were all calcu lated in the usual manner. Physiologic dead space (Vp), alveolar ventilation (Va), and the alveolar 02 pressure (Pao2) were calculated from the alveolar equation, substituting PeToo- for Paco,. Alveolar CO pressure (Pac0), the steady state diffusing capacity (DlSS), and the fraction of CO removed from alveolar ventilation (FCO) were calculated ac cording to Filley et al.11 The single breath diffusing capacity (DlSB) was measured in duplicate according to Forster et al11 with a special instrument13 The original procedure was modified by using an alveolar volume (Va) calculated from the single breath helium dilution ratio14 instead of Va calculated by addition of the inspired volume to a residual volume measured separately by a rebreathing technique.13 Each subject had a dry run with air; the inspiratory circuit was then flush*! with balloon gas. A buzzer signaled the ten-sec ond breath-holding period, and an interval of a* least five minutes was allowed between dupli cate studies. When the inspired volume *** less than 90% of the FVC, the procedure repeated. A predicted DlSB was calculat`d from regression equations derived from *1*' normal subjects.1#'13 Arch Environ Health--Vol 25, Oct 1972 I BB 0015834 | LOW EXPC&RE TO ASBESTOS-MURPHY ET AL 255 Ota A/udy*"*--Carbon monoxide and C02 measured with nondispersion infrared .tronleters which were calibrated with four of known composition before and after ,ih run. Oxygen was analyzed with a paranrtgnetic instrument, and He with a tempera>lire compensated thermistor catharometer. 1'hese linear instrumenta were calibrated with nitrogen and one gas of known concentration lefore and after each run. Switching of gases, valves, sample pumps, and meters was accomplished with a solenoidcontrolled instrument designed for this survey so that one individual could control the entire procedure. Carboxyhemoglobin (COHb) saturation and CO "back pressure" were estimated by breath holding17 before and after each test in ten randomly selected nonsmokers and ten smok ers. Initial COHb saturations ranged from 1% to 2% in nonsmokers and up to 6% in heavy smokers in the afternoon. Each DlSB, steady state diffusing capacity at rest and exercise (DlSS-R and DlSS-Ex) raised COHb satu ration by about 1%, 2%, and 4% respective ly. Neglect of back pressure may have re sulted in an underestimation of Dl of up to 4% in some cases. Results Respiratory Function of Pipe Coverers and Controls.--The two groups were vir tually identical with respect to age, stature, present smoking, and pack-years of cigarette consumption (Table 1). "Restrictive" im pairment was more common in the pipe coverers in that their FVC, FEVl and total lung capacity (TLCH) was significantly lower (P < .01). From the measurements cited in Table 1, there was no evidence that obstructive air ways disease was more common in the pipe coverers than in the controls. Their mean FEV,% was 72.1 8.9 compared to controls with 73.7 8.1%; there was no significant difference in airways resistance, specific con ductance, or functional residual capacity. In Fig 1, the subjects are arranged in order of their FEV^; nine pipe coverers (10.7%) and five controls (7.1%) had an FEVj% which was more than 2 standard deviations below the predicted mean, an in significant difference. The presence of clini cal asbestosis, shown in relation to FEV^, was not particularly related to obstructive lung disease; only one such pipe coverer had a significantly reduced FEVi%. A reduction of Dl may be caused by obstructive air ways disease, particularly with emphysema.7 However, only four pipe coverers and two controls had a significant reduction of both Dl and FEVj%. Among the tests of respiratory gas ex change, the largest difference between pipe coverers and controls was found concerning DlSB and DlSS-Ex (P < .01). Not signifi cantly different were DlSS-R and Fco-Ex. Discriminant function analysis confirmed these observations. Definition of a Normal Control Group.-- The control group was selected to match the pipe coverers solely by age and duration of employment, and therefore it included work ers with significant cardiopulmonary impair ment. It was thought that this represented the expected background of nonasbestos re lated disease prevalent in men working in the same locality for the same number of years in related occupations.1 However, it was also of interest to compare the pipe coverers to individuals without known cardio respiratory disease because, for several of our tests, regression equations for normal performance are not available. Therefore, we selected a "normal" control group by excluding from the 70 controls 30 persons who had symptoms or a history of cardiore spiratory illness with confirmatory evidence by physical examination, chest roentgeno gram or both (Table 3). In Table 1, we compared the pipe coverers not only to all controls but also to those 40 normal controls. The latter were about three years younger and smoked slightly less, but these differences were not significant. Pre dictably, the respiratory function of the pipe coverers differed in the same respects from the normal controls as from all controls. In addition, the new comparison showed the pipe coverers to have significantly increased Petc02 and VeO,, during exercise, and sig nificantly reduced DlSS-R, diffusing con stant K, and Fco both at rest and during exercise. A number of other variables was virtually the same for all three groups in cluding Vo2, Vco2, f, Vt and a residual vol ume obtained by subtracting FVC from the total lung capacity calculated from the single breath He dilution. Arch Environ Health--Vol 25, Oct 1972 f_BB 0015835 J -_________________T Table 1.-- Respiratory Fi/nction \)f eo Normal Controls All 70 Centrals 1. Stature and smoking Age, yr Weight, kg (lb) Height, cm (in) Body surface area, sq m, (SSA) Smoking, pack years 2. Mechanics and lung volumes Vital capacity, liter (FVC) Timed FVC. liter in 1 sec (FEVi) (FEV,/FVC) X 100 = FEV,% Airway resistance, cm/liter/sec (Ra) Specific conductance, liter/see, em/liter (GA/Vt) Functional residual capacity (plethysmography), liter (FRCbp) Total lung capacity, (He) liter (TLCh) 3. Steady state studies Ventilation, liter/min/sq m (Ve) rest Ventilation, liter/min/sq m (VE) exercise 02 uptake, ml/min/sq m (Vot) rest O; uptake, ml/min/sq m (vo,) exercise Ventilatory equivalent for Oi, liter/100 ml (ileOi) rest Ventilatory equivalent for Ot. liter/100 ml (VeOi) exercise Respiratory exchange ratio (R) rest Respiratory exchange ratio (R) exercise End'tidal CO* tension, mm Hg (PETco*) rest End-tidal CO, tension, mm Hg (PETcoj) exercise Physiologic dead space, ml (Vo) rest Physiologic dead space, ml (Vo) exercise Fraction CO removed, % (FCO) rest Fraction CO removed. % (Fco) exercise Diffusing capacity, mi/min/mm Hg (DtSS) rest Di",is-n- capacity, ml/min/mm Hg (DlSS) exercise 4. Single breath Diffusing capacity, ml/min/mm Hg (DlSB) Krogh's diffusion constant (K) Mean 40.3 77 (171) 175(68.7) 1,92 17,2 4.60 3.47 75.5 1.97 0.191 2.96 5.86 5.38 16.31 166 667 3.26 2.45 0.85 0.87 37.5 43.2 209 423 46.5 39.5 18.8 30.4 36.8 4.6 SD Mean 9.9 13 (28) 6(2.5) 0.17 16.5 44.0 78(174) 173 (68.0) 1.92 25.4 0.58 0.47 6.1 0.69 0.047 0.59 0.74 4.28 3,15 73.7 2.11 0.183 2.94 5.59 1.00 3.32 22 120 0.52 0.26 0.10 0.05 3.0 3.3 49 117 5.7 4.9 5.7 6.1 5.59 17.06 165 662 3.40 2.59 0.85 0.88 36.5 41.9 208 421 44,8 37.4 17.3 28.6 6.0 34.3 0.8 4.5 * Pipe coverer* significantly different (P < .01) from normal controls only, t Pipe coverers significantly different (P < .01) from normal controls and all controls, t Pipe coverers significantly different (P < .01) from all controls only. SD 10.3 13 (28) 6(2.5) 0.16 24.4 0.78 0.65 8.1 0.75 0,054 0.73 0.86 1.43 3.53 22 105 0.72 '0.43 0.10 0.07 3.8 3.7 52 108 7.2 6.0 5.6 6.4' 6.3 0.8 In the subsequent analysis of results and discussion, this special group of normal con trols was not further considered. Relationship of Clinical "Asbestosis" to Abnormalities of Respiratory Gas Exchange. --Figure 2 illustrates that mean values for the three Dl tests decreased with an in creasing number of clinical abnormalities; even the pipe coverers with none or only one of the clinical criteria had lower Dl values than the controls. Individual clinical find ings also correlated with tests of gas ex change; radiologic abnormalities correspond ed most closely; even the readings of ques tionable (code 3 or UICC 0/1) and slight (code 4 or UICC 1/1) asbestosis had mean ing in statistical terms while they had been of little importance in individual cases. Clubbing was least related, reflecting per haps the nonpulmonary causes of this ab normality or the difficulties in accurate as sessment. A principal objective of the Dl and Fco studies was to discover individuals with im paired lung function, perhaps due to asbes tosis, who were considered "borderline" or "negative" by our clinical criteria. For this purpose we defined abnormal values as those that were more than 2 standard devia tions below the mean of all controls. Figure 3 shows that workers with a "positive" clini cal diagnosis also had abnormal DlSB, DlSS, and Fco at rest and during exercise, or were unable to exercise. Of greater interest were workers with im paired gas exchange and only two positive clinical findings. Three such pipe coverers had no abnormalities suggesting any other diagnosis than asbestosis, while among the three controls, two had bullous emphysema, and one a lobectomy for carcinoma. This suggests that inclusion of diffusion tests among the criteria for diagnosis might define additional individuals with lung disease. There were eight pipe coverers with im paired Dl and Fco who had no clinical Arch Environ Health--Vol 25, Oct 1972 7 BB 00?5836_7 o ------ __ --------------------------------------84 Pips Coverers | Mean SO I 43.2 I 76 (168) I 171 (67.4) I 1.88 1 24.8 11.0 10 (23) 6(2.2) 0.14 20.2 1 3.87 1 2.78 1 72.1 I 1.96 I 0.213 1 2.77 1 5.15 0.70 0.59 8.9 0.73 0.069 0.66 0.90 I 5.74 I 17.72 170 1 668 3.36 1 2.66 I 0.83 0.87 36.4 40.9 222 394 42.8 35.5 15.9 25.8 1.37 3.86 25 98 ' 0.62 0.47 0.12 0.07 3.6 4.4 53 115 6.7 6.9 4.9 5.8 30.5 4.2 6.6 0.8 abnormalities and a normal FVC. Only follow-up ex aminations will show whether such isolated im pairment of gas exchange has meaning in terms of de tecting early asbestosis.20 There is support for this possibility. Four of the pipe coverers with isolated Dl reduction were restudied three years later. At that time, one had developed dry rales and dyspnea, one had finger clubbing, and one had emphysematous bullae and a markedly reduced FEVt%. Follow-up Studies.--In 1970, a third visit was made to the shipyard, 3y2 years af ter the second survey. Eight surviving workers with as Significance t t : t * * t t * bestosis and 24 other pipe xbverers were re evaluated. Among the latter, two ad ditional individuals now fitted our clinical criteria for asbestosis. The mean FVC of these 32 workers had decreased from 3.70 to 3.53 liters or 4.5%. This is significant (F < .01), but amounts to only 47 ml/yr, com pared to a 20 to 100 ml/yr reduction predict ed for aging alone.18-1 The mean DlSB had decreased much more, from 26.9 to 23.5 ml/min/mm Hg, or 12.6%; this was 0.96 ml/ yr, nearly four times the predicted decrease of 0.26 ml/yr.20 The number of workers with a reduced FVC had increased from 12 to 14, and those with reduced DlSB from 12 to 19 (Fig 4). None of those who were abnormal in these respects in 1966 had improved. Correlations Among Physiologic Tests for All Subjects.--Many tests correlated signifi cantly with others (P < .01 if r > .21) be cause they reflected the same type or location of functional impairment (Table 3). In some instances, close correlations suggested simplifications for future surveys.21 Forced vital capacity corresponded well to the max imal inspiratory effort preceding the DlSB test (r - .94), an observation that indicates adequate performance by most subjects. Cor relation of FEVX% with airways resistance and with specific conductance was poor (r = .09 and .37), because in this study nearly all values fell within the broad range of normal (Fig 1), whereas previous studies showing good correlations contained many patients Table 2.--"Normal" Controls and Controls Cardiopulmonary Disease Normal controls i. Asymptomatic, or no more than 1+ dyspnea, 1+ cough, or 1-|- sputum with negative history 2. Positive history but without symptoms and without residual verified by physical examination and roentgenogram Total 40 Controli with cardiopulmonary disease 1. Cardiac Acquired heart disease, symptomatic Congenital heart disease, symptomatic 2. Pulmonary obstructive Chronic bronchitis with obstructive syndrome Bullous emphysema with or without obstruction Bronchial asthma with signs and symptoms Emphysema and mid thigh amputation 3. Pulmonary restrictive Healed tuberculosis with radiologic residua Marked pleural thickening, old empyema Chest trauma with marked radiologic residua Lobectomy for bronchogenic carcinoma Pneumonectomy for pulmonary arteriovenous aneurysm "Diffuse Interstitial fibrosis" from smoke inhalation (asbestosis by our clinical criteria) Total 4 2 5 5 3 1 3 2 2 1 1 1 30 Arch Environ Health--Vo! 25, Oct 1972 I BB 0015837 1 258 LOW EXPOSURE TO ASBESTOS^ftURPHY ET AL Table 3.--Correlation Coefficients (r) Among Physiologic Tests* FVC* liter DLSS-R OLSS-Ex FCO-R FCO-Ex Vc-R VE-Ex PeTcoi-R PetcOj-E* Maximal 1C preceding DlSB maneuver Total lunR volume from SB He dilution FVC 0.94 0.74 DlSS Rest 0.39 --0.09 0.37 OLSS Exercise 0.52 0.76 -0.23 0.49 DLSB 0.52 0.57 0.72 0.51 0.61 FCO Rest 0.68 0.61 "T ----~ -0.63 0.53 ' * For J54 (pipe coverers and controls combined) subjects: any r > .16: P < .05; and r > .21: P < CONTROLS Number PIPE COVERERS Number Fig 1.--Pipe coverers and controls arrange', . ing to decreasing FEV,%. Subjects with a.* * ^ bestosis are indicated in black. The frequent bution of FEV,% was about the same for both-, asbestosis was not significantly related to . rev,*. Arch Environ Health--Vol 25, Oet 1972 1 BB 0015838 1 LOW EXPOSURE TO ASBESTOS--MURPHY ET A 259 Vith severe obstructive dis ease. Concerning the relation ship between FVC and Dl, it has been said that DlSS is more affected by ventila tion-perfusion discrepancies, while DlSB is related more to lung volume; neverthe less, here FVC related equally well with both (Ta ble 3). All three Dl measure ments were closely related, particularly DlSS-R with DlSS-Ex (r *= ,75) and DlSS-Ex with DlSB (r -- ,72). The scattergram of Fig 5 shows that all indi viduals with an abnormal DlSS-Ex also had an ab normally low DlSB, but the reverse was not true. There was a good correla tion between Fco and DlSS (Fig 6) but with considerable scatter. This was largely because Fco was inversely related to minute ventilation (r * -.63 at rest, -.76 during exercise), while DlSS was only slight ly affected by ventilation (r * .09 and .23, respective ly) (Table 3). As a corol lary, Petco, was much more related to Fco than to DlSS (Table 3). In other words, Fco was very sensitive to voluntary or pathologic hy perventilation while DlSS was much less affected by this. X-Ray 5B6 Asbestosis 1-4-1.--J--.1--2m Dyspnea 2+ Dry Roles Clubbing .__ frKTffll VC <80% B 0 Fco Rest 12SD^ Fco Ex 4 2SD Dl Rest <10 Ex 4 2SD Dl SB 4 2SD X-Ray 596 Dyspnea 2+ Dry Rales Clubbing VC <80% Fco Rest 4 2SD FcoEx <2SD Dw Rest < 10 0k Ex 42SD Dl SB 42SD Fig 3.--Pipe coverers and controls arranged from left to right accord ing to the number of positive clinical criteria; three or more signified asbestosis. Abnormalities of gas exchange are also shown. U signifies un able to exercise; N indicates normal controls (Table 2). Fig 4.--Comparison of FVC and DLSB in 32 pipe coverers studied three years later. Eight had asbestosis in 1966, and ten in 1970. Solid dots indicate performance in 1966 and arrow for 1970. The tines indicate 2 standard deviations down from the mean of controls. The number of work ers with a significantly decreased FVC increased from 12 to 14, and those with a reduced DLSB from 12 to 19 during the interval. 32 PIPE COVERERS Comment Respiratory abnormalities in advanced asbestosis have been well described. Reduc tion of lung compliance1-- -5 is reflected by restriction of Arch Em iron Health--Val 25, Oct 1972 1~Bb"0015839~7 veolar-capillary block" w* used to characterize the prl'l cipal impairment resultii^1 from asbestosis.2``M3.23'33.3u- It is now recognized that tli concept of alveolar-capillar^ block is an oversimplification^ because ventilation-perfusio discrepancies contribute, may be largely responsible, for the A-a 02 gradient. In, asbestosis, uneven ventila-, tion and perfusion have beenv demonstrated whenever re fined techniques were; used.s>;9>3i In such a "mixed impairment of CL transport mechanism,"5 the theoretical and practical difficulties of separating "diffusion gradi ents" from "venous admix ture gradients" are consid erable.21 At any rate, neither the pathologic changes nor the physiologic and clinical alterations of asbestosis dif fer from those associated * Pipe Coverers 0 Normal Controls + Other Control with "usual interstitial pneu monitis."34 Fig 5.--Correlation between steady state exercise (DLSS-Ex) and breath holding (DLSB) diffusing capacities. Broken lines indicate 2 standard de viations down from the mean for all controls. The Physiologic Altera tions of Early Asbestosis.-- Definition of such changes would be of greatest interest vital capacity.1-8-22-29 This has been attribu for monitoring of exposed workers. Wright2 ted to pulmonary fibrosis while the added ef first reported functional abnormalities in as fects of pleural complications have been con sidered more recently.8'24'30-31 Hyperventila tion, especially during exercise, is common.3'4'8'24-20'2' Wright2 has pointed out that bestos workers without clinical or radiograph ic manifestations and such cases have been recognized subsequently.3-7-22'30'32'35 Inter pretation of these abnormalities is difficult dyspnea in asbestosis is more the result of an because they may not be caused by asbes increased ventilatory requirement than of a tos.38 Among our pipe coverers without clini reduced ventilatory capacity. Such hyperven cal or radiologic signs of asbestosis, IS had tilation, stimulated by slight hypoxemia and one or more physiologic alterations (Fig 3). increased lung and chest wall reflexes, has The discriminant value of function tests been thought to occur early, but this was not has been evaluated by comparing cases of borne out by our studies (Table 1). asbestosis certified by the British Pneumo Impaired CL exchange was first inferred coniosis Medical Board with workers in as from "oxygen debt," reduced ventilation on bestos-using industries without certifiable CL breathing28 and oximeter studies28-22 and disease.3 7'23 38 Even though the "diagnosis eventually was confirmed by demonstration of the Medical Board is reliable,"38 the of increased A-a CL gradients.2'513-29'31 This significance of impaired function in the "alveolar-respiratory insufficiency" has been absence of clinical evidence of disease ob nscrihed to alterations in the thickness and viously cannot be evaluated in this way. size of the pulmonary membrane.2'28 28 With Few serial examinations are available. Lea- demonstration of a reduced Dl,13 the term "al thart38 selected 12 loggers initially free of Arch Environ Health--Vol 25, Oet 1972 1 BB 0015840 | disease;, over a three- to 25 r nine-year period, five showed a'drop of DlSS to less than 80% of their original read ing. Among these, three de veloped rales, one emphyse to 03 20 ma, and one convincing evidence of asbestosis. More persuasive data come from Hunt.0 He followed 36 work ers with a DlSB below 60% but without clinical signs of asbestosis for four to five years. During this time, four died of neoplasia and had asbestosis at death, and 14 others developed clinical and I 15 10 c .V. O m ov*m - O 4 O* O >.'oV y*. * M* .4 radiological signs of asbes <0 tosis. Our own three-year Oj4 5 follow-up included 24 pipe r=.8l coverers with fewer than three clinical criteria for as bestosis; among those with initially reduced FVC and 20 30 40 Fco Ex, % 50 DlSB as the only abnor mality, two developed dysp nea, dry rales, and club Fig S.--Correlation between steady state exercise (DLSS-Ex) dif fusing capacity per square millimeter of body surface area and fraction of carbon monoxide removed (FCO-Ex). The code is the same as in Fig 5. bing, and one emphysema tous bullae. The best estimate of disease prevalence in The question of sensitivity of individual this group with uniform exposure, in our function tests has been considered. Williams opinion, was the duration of exposure. Ac and Hugh-Jones3 and Leathart4 suspected cordingly, we calculated correlation coeffi that Dl might be the most sensitive index of cients between this variable and our clinical impairment; and it is now recognized that criteria as well as selected tests of re reduced Dl may be found in exposed indi spiratory gas exchange. Pipe coverers and viduals who have normal FVC.0'7 23'31'32-30 controls were combined for these analyses. Becklake and associates,* on the contrary, Significant correlations were found for concluded from a population-oriented study DlSB (-0.395), VC (-0.386), x-ray that reduced FVC and inspiratory capacity (0.356), dyspnea (0.320), and DlSS-Ex (IC) are the earliest signs of asbestosis. (-0.318) (Table 4). Discriminant function Their data are difficult to interpret because (multiple r-) for the clinical variables ex they chose as controls asbestos workers with cluding VC was 0.246; adding VC it was out radiographic abnormalities. Also, their 0.299 and increased to a maximum of 0.311 series purposely was weighted with older in when DlSB was the only test of respiratory dividuals*; indeed, before these older indi gas exchange added to the clinical variables viduals were included, both FVC and Dl (Table 4). These associations are not as were significantly reduced in early disease.37 high as those reported by Regan et aF4 most In our series, the mean Dl was reduced even likely because our inclusion of controls tend among workers with roentgenograms reveal ed to lower these values. However, we be ing questionable asbestosis. Prevalence sur lieve that this represents a more accurate veys may be misleading on the question of assessment of the true association. which type of examination gives the earliest Chronic Obstructive Lung Disease.--Ir. evidence of impaired function. The answer early studies of severe asbestosis, em can come only by following a defined popu physema and chronic bronchitis were lation over a period of years.8 Unfortunately, prominent28-2*; later, obstructive lung dis no such studies have been reported. ease with reduced maximal breathing capac- Arch Environ Health--Vol 25, Oct 1972 | BB 0015341 1 V 262 rLOW EXPOSURE TO ASBESTOS--MURPI1Y ET AL Table 4.--Simple Correlation Coefficients and Discriminant Function Analysis for Clinical Variables and Selected Tests of Pulmonary Function vs Duration of Exposure*______ 1. Simple correlation coefficients vs duration of exposure Age Dyspnea Paronychial angle Rales VC X-ray DlSS-R FCO-R dlSS-Ex FCO-Ex DlSB +.333 + .320 +.086 +.229 -.386 +.356 -.175 --.203 -.318 -.309 -.395 physiologic data (Fig 1); moreover the prevalence of obstructive impairment was no greater among pipe coverers with asbestosis than among those without clinical or radiologic manifestations of the disease. Critique of Methods.-- Determination of DlSS re quires knowledge of the ef 2. Discriminant function Clinical variables (dyspnea, rales, paronychial angle, and x-ray films) Without VC With VC Clinical variables with VC plus () DlSS-R () FCO-R (c) DlSS-Ex <rf) FCO-Ex (e) DLSB Multiple r1 .246 .299 .299 .301 .301 .299 .311 fective alveolar CO partial pressure, Paco. Usually, this is calculated with the alveo lar equation from arterial and expired Pco,.11 To ob viate arterial pfincture, Leathart30 used a rebreathing technique to estimate arte rial PcOo. This was done as * Pipe coverers and controls combined. a separate procedure and without steady state and ity and FEVi% and increased Ra and re caused poor-reproducibility. Bates et al41 and sidual volume (RV) was demonstrated in others8 have used end-tidal sampling for CO, up to one half of patients.4-21-31*33 Recent while we have preferred to sample Petco, be observers, however, have thought that this is cause the latter can be done directly while the not an important complication.2-3-23-20-32 Fur much larger sample needed for CO analysis thermore, such tests as the FEV^ and requires a special and capricious end-tidal peak flow may be misleading31 because, with sampler. In 102 normal volunteers studied in increased elastic recoil of pulmonary fibro our laboratories, Petco,, averaged 2.1 mm less sis, this value may be larger than normal.8 than Pac0,, at rest, and 2.9 mm less during Most recently, it has been suggested on exercise. For eight asbestotics studied by physiologic grounds that early asbestosis both methods, the arterial-end-tidal COs dif may result in obstructive disease of small ference was similar: Paco2 averaged 37.5 airways.39 However, this is not borne out by mm and PetC02 was 35.4 mm; while dur our pathologic studies of lung biopsies in ing exercise, these means were 33.2 and 35.3 early disease.40 mm, respectively. We concluded that our Chronic bronchitis and emphysema are common particularly among older male smok ers; therefore, comparison of large groups of exposed workers must be made with controls living in the same area, matched for sex, age, duration of employment, and smoking habits. method, compared with the original Filley technique,11 resulted in slightly lower values for DlSS and tended to accentuate the dif ferences between normal subjects and those with lung disease. The same may apply to end-tidal CO sampling. Only three surveys have included con For calculation of DlSB, we used an trols,8-32-3! and actual data were shown only alveolar volume, Va, calculated from single by Kleinfeld et al,32 who compared 56 asbes breath He dilution instead of a volume ob tos workers with 50 men without dust expo tained separately by He rebreathing. Aside sure. They found more individuals with re from saving much time, this has been advo duced FEVi% and increased RV in the cated as a better method because the distri exposed group, but it contained twice as bution of He during breath-holding is said many smokers. In our clinical study, chronic to be more relevant to the "effective Va" to obstructive lung disease was equally common which CO is distributed during the same among exposed workers and controls,1 and deep breath.14 In normal subjects, the two Arch Environ Health--Vol 25, Oct 1972 c 1 BB 0015842 1 LOW EXPOSURE TO ASBESTOS--MURPHY ET AL 263 volumes should be the same. Actually, in 102 normal subjects in our laboratory, the ny*n rebreatliing Va of 4.19 liters (STFD) was virtually the same as the single breath Va of 4.48 liters; and in seven of the pipe coverers with asbestosis, the mean Dl calcu lated from conventional rebreathing Va was 17.1 ml/min/mm Hg while the mean from single breath He was 19.0 ml. With uneven lungs, as in emphysema, the single breath Va may be smaller than the rebreathing Va,7-14 and hence the Dl would be under estimated by the simplified technique. Again, this would tend to exaggerate the differences between normal and abnormal subjects. Conclusions Tests of respiratory gas exchange gave corroborative evidence that some pipe cover ers had asbestosis despite the low level of their exposure. All workers with three or more clinical criteria of asbestosis had mark edly reduced Dl and Fco. In addition, sev eral with only two clinical criteria also had impaired gas transfer suggesting that inclusion of such data might improve our diagnostic acumen. Among the several tests employed, DlSS-Ex and DlSB were reliable and re producible methods of assessing interstitial lung disease, while DlSS-R and particular ly Fco-R were sensitive to minor variations in minute ventilation. A unique feature of this study as a de tailed analysis of the cardiopulmonary sta tus of controls selected on epidemiologic grounds. Although their mean FVC was nor mal, 30 of the 70 had cardiopulmonary ab normalities. This suggests that surveys look ing at a single or only a few factors may be misleading. In spite of these abnormalities among the controls, the pipe coverers had significantly lower FVC, FEV,, TLC, DlSS-Ex, and DlSB (PC.01) indicating the major effect of interstitial lung disease on these measurements. The presence of an equal number of cigarette smokers in the two groups, as well as their equal FEV,%, argues for the overriding importance of the dust exposure in producing the manifesta tions of lung disease herein described. The earliest changes of interstitial pneu monitis, consisting of cellular infiltration and regenerating or metaplastic epithelium, probably cannot be detected by any meth od other than tissue examination.30-*1'3*'40 Possibly, characteristic dry rales may be heard at this time.38 Decrease of Dl and FVC cannot be appreciated until later be cause of the large standard error of normal range and because of the large pulmonary reserve. The geometric limitations of radiographic techniques are such that interstitial pneumonitis cannot be detected with some certainty until the disease is moderately advanced34-40; dyspnea, finger clubbing, and cyanosis are late manifestations. No definitive studies provide an answer to the relative importance of individual criteria in detecting early asbestosis. Our results (Fig 3) support the impression of Thomson et al7 that in some patients restriction of lung volume is the earliest manifestation of disease while in others abnormalities of gas exchange appear earlier. In our longitudinal studies, DlSB deteriorated more rapidly than VC, suggesting that this measurement may prove more sensitive. Asbestosis is a serious disease which is irreversible when advanced. It can occur with exposure to low levels of dust, and such exposure is increasingly common. Informa tion on the diagnosis and early detection of asbestosis therefore is important and timely. This investigation was supported in part by Pub lic Health Service research grant HE-05933 from the National Heart and Lung Institute, training grant HE-5562, and Career Award HE-1173. Homayoun Kamezi, MD, supplied data from the Pulmonary Laboratory of Massachusetts General Hospital. The solenoid-controlled instrument designed for this survey was a Collins Modular Clinical Spiro meter P-1280. Warren E. Collins Co. constructed and lent spe cial equipment, technical assistance, and provided transportation. The Hewlett Packard Co. also lent equipment. References 1. Murphy RLH Jr, Ferris BG Jr, Burgess WA. et al: Effects of low concentrations of asbestos: Observations in shipyard pipe coverers. New Eng J Med 285:1271-1278, 1971. 2. Wright GW: Functional abnormalities of in dustrial pulmonary fibrosis. Arch Indutir Health 11:196-203,1955. 3. Williams R, Hugh-Jones P: The significance of lung function changes in asbestosis. Thorax 15:109119, I960. 4. Leathart GL: Clinical, bronchographic, radio logical and physiological observations in ten cases of asbestosis. Brit J Induetr Med 17:213-227, 1960. 5. Bjure J, Sdderholm B, Widimsky J: Cardiopul- Arch Environ Health--Vol 25, Oct 1972 J BB 0015843 | 264 monry function studies in workers dealing with asbestos and glasswool. Thorax 19:22-27, 1964. 6. Hunt R: Routine lung function studies on 830 employees in an asbestos processing factory. Ann NY Acad Sci 132:406-420. 1965. 7. Thomson ML, McGrath MW, Smither WJ, et al; Some anomalies in the measurement of pulmon ary diffusion in asbestosis and chronic bronchitis with emphysema. Clin Sci 21:1-13, 1961. 8. Becklake MR, Foumier-Massey G, McDonald JC, et al: Lung function in relation to chest radiographic changes in Quebec asbestos workers. Bull Physio-Path Reap 6:637-659. 1970. 9. Bohlig H, Bristol LJ, Cartier PH, et al: UfCC/Cincinnati classification of the radiographic appearances of pneumoconioses: A co-operative Study by UICC committee. Chrst 58:57-67, 1970. 10. Dubois AB, Botelho SY, Bedell GN, et al: A rapid plethvsmographic method for measuring tho racic gas volume. J Clin Invest 35:322-326, 1956. 11. Filley GF, Macintosh DJ, Wright GW: Car bon monoxide uptake and pulmonary diffusing ca pacity in normal subjects at rest and during exer cise. J Clin Invest 33:530-539, 1954. 12. Forster RE, Fowler WS, Bates DV, et al: The absorption of carbon monoxide by the lungs during breathholding. J Clin Invest 33:1135-1145, 1954. 13. Marks A, Cugell DW, Cadigan JB, et al: Clinical determination of diffusing capacity of lungs: Comparison of methods in normal subjects and pa lienla wall "alveolar-capillary block" syn drome. Amer J Med 22:51-73, 1957. 14. McGrath MW, Thomson ML: The effect of age, body size and lung volume change on alveolar capillary permeability and diffusing capacity in man. J Physiol 146:572-582, 1959. 15. Cadigan JB, Marks A, Ellicott MF, et al: An analysis of factors affecting the measurement of pulmonary diffusing capacity by the single breath method, j Clin Invest 40:1495-1514, 196L 16. Gaensler EA, Wright GW: Evaluation of re spiratory impairment Arch Environ Health 12:146189,1966. 17. Jones RH, Ellicott MF, Cadigan JB, et al: The relationship between alveolar and blood carbon monoxide concentrations during breath holding. J Lab Clin Med 51:553-564, 1958. 18. Cara M: Depouillement de l'enquete statistique sur les valeurs fonctionnelles ventilatoires: Etude preliminaire, in Copee G, Houberechts A, Lehmann G. et al (eds): Arbeitsphysiologische und arbeitspatholngische Studien, Schriflenrethe: Arbeitshygiene und Arbeitsmedizin, Nr. I Luxemburg, Europaische Gemeinschaft fur Kohle und Stahl, Hohe Behorde, 1961. pp 101-103. 19. Kory RC, Callahan R Boren HG, et al: The Veterans Administration-army cooperative study of pulmonary function. Amer J Med 30:243-258, 1961. 20. Cotes JE, Hall AM: The transfer factor for the lung: Normal values in adults, in Arcangeli P fed): Normal Values for Respiratory Function in Man. Torino, Italy, Panminerva Medica, 1970, pp. 327-343. 21. Gaensler EA, Smith AA: Lung function mini computer. Presented aa a scientific exhibit, annual meeting of the American Thoracic Society. Clev land. May 25-27, 1970. 22. Bader ME, Bader RA, Tierstein AS, et 3 Pulmonary Function in asbestosis. Ann NY Aca Sci 132:391-405, 1965. 23. Thomson ML, Pelzer A-M, Smither WJ: Tb discriminant value of pulmonaiy function tests 1 asbestosis. Ann NY A cud Sci 132:421-136. 1965. i 24. Hany A. Burckhardt P, Buhlmann A: Zu Klinik und Pathophysiologic der Lungenasbestos* Schweiz Med Wschr 97:597-603, 1967. > 25. Ardalan P: Atemmechanische und elektrokat diographische Untersuchungen bei Lungcnfibroso Prax Pneumol 22:781-792, 1968. ; 26. Gemex-Rieux C. Balgairics E, Claeys C: Coni siderations sur les troubles respiratoires dans I'as. bestose. J Franc Med Chir Tharac 8:193-198, 1954. 27. Bastenier H, Denolin H. De Coster A, et al: Etude de la fonction respiratoire dans l'asbestosc pulmonalre. Arch Mai Profess 16:546-553, 1955. 28. Roetnheld L, Kerapf H, Wedler HW: Unier- suchungen fiber die Lungenfunktion bei Asbestos*. Dcutsch Arch Klin Med 186:53-70, 1940. 29. Bader ME, Bader RA, Sclikoff IJ: Pulmonary function in asbestosis of the lung: An alveolar-capil lary block syndrome. Amer J Med 30:235-242, 1961. 30. Gaensler EA, Kaplan Al: Asbestos pleural effusion. Ann Intern Med 74:178-191, 1971. 31. Mattson S-B, Ringqvist T: . Pleural plaques and exposure to asbestos. Scand J Reap Dis, suppi 75, pp 1-40. 1970. 32. Kleinfeld M, Messite J, Kooyman O, et ai: Effect of asbestos dust inhalation on lung function. Arch Environ Health 12:741-746. 1966. 33. Read J, Williams RS: Pulmonary ventilation, blood flow relationships in interstitial disease of the lungs. Amer J Med 27:545-550, 1959. 54. Gaensler EA, Carrington CB, Coutu KE: Chronic interstitial pneumonitis. Clin Notes Reap Dis 10:3-16, 1972. 35. Gandcvia B: Pulmonary function in asbestos workers. Amer Rev Reap Dis 96:420-427, 1967. 36. Leathart GL: Pulmonary function testa in asbestos workers. Trans Sac Occup Med 18:49-55, 1968. 37. Becklake MR, Foumier-Massey G, McDonald JC, et al: Relationship of functional to radiographic changes in Quebec asbestos workers, in Proceedings of the Second International Conference on Biologi cal Effects of Asbestos, Dresden, 1968. 38. Regan GM, Tagg B, Walford J, et al: The relative importance of clinical, radiological and pul monary function variables in evaluating asbestosis and chronic obstructive airway disease in asbestos workers. Clin Sci 41:569-582, 1971. 39. Jodoin G. Gibbs GW, Macklem PT, at al: Early effects of asbestos exposure on lung function. Amer Rev Reap Dis 104:525-533,1971. 40. Gaensler EA, Carrington CB, Coutu RE, ct al: Pathologic-physiologic-radiologic correlations in pneumoconioses. Ann NY Acad Sci, to be published. 41. Bates DV, Woolf CR, Paul GI: Chronic bron chitis: A report on the first two stages of the coordinated study of chronic bronchitis in the De partment of Veterans Affairs, Canada. Med Serv J Canada 18:211-303,1962. Arch Environ Health--Vol 25, Oct 1972 I 4k-B--B--- 0015344 *I