Reduced Diffusing Capacity as an Isolated Finding in Asbestos- and Silica-Exposed Workers

Reduced Diffusing Capacity as an Isolated Finding in Asbestos- and Silica-Exposed Workers* joe G.N. Garcia, M.D., F.C .C.P.; David E. Griffith, M.D., F.C.C.P.; james S. Williams, M.S.; WilliamJ Blevins; and Richard S. Kronenberg, M.D., F. C . C . P.

From a cohort of 286 patients referred to an Occupational Medicine Clinic because of exposure to asbestos and/or silica, we identified 53 patients with a reduced diffusing capacity (Dco) ( <75 percent predicted) as their only abnormality. Specifically, their clinical evaluation, chest roentgenograms, and remaining pulmonary function test results were all normal. These patients were divided into nonsmokers (n=l3) and smokers (n=40). The significance of the isolated reduction in diffusing capacity in these patients (n=53) was explored with graded exercise testing (n=l9) and bronchoalveolar lavage (BAL) (n =50). The results obtained from the patients with reduced diffusion were compared with those obtained from comparable smoking (n=35) and nonsmoking patients (n=37) in the original cohort who had normal chest roentgenograms and normal results of pulmonary function studies, including normal Dco values (2:75 percent of predicted value). Patients with low diffusion demonstrated a tendency for elevated alveolar to arterial 0 1 differences both at rest and during exercise, and a significant reduction in exercise capacity (Vo, max) was observed in the smoking patients with reduced diffusion when compared with their smoking counterparts with normal diffusion. All other exercise testing indexes were normal in the study groups and there was no correlation between the percent predicted Dco value and any of the exercise variables. In contrast, BAL revealed significant differences between patient groups. Both the smoking and nonsmoking patient groups with low Dco values had greater

numbers of total BAL cells, alveolar macrophages, neutrophils, lymphocytes, and eosinophils in their BAL fluid than did their comparable controls with normal diffusion values. These differences were statistically significant (p<.05) for total BAL cells and total macrophages in the nonsmoking patients and for total BAL cells, total macrophages, and total lymphocytes in the smoking patients expressed as either the total cell number per BAL or total cells per milliliter of BAL. In contrast to the observed exercise testing results, there was significant and inverse correlation between Dco values and each BAL cell type for all four groups combined as well as nonsmokers alone. The Dco values from smokers were significantly and inversely correlated with total BAL cells and total macrophages. These results suggest that the finding of a reduced Dco may be related to an active inftammatory process in the lung caused by occupational dust exposure. Patients with both dust exposure and low Doo values should be followed up closely even if the low Dco value is an isolated finding; however, the utility of noninvasive exercise testing in the evaluation of these patients requires further investigation.

The diagnosis of pneumoconiosis, even in the appropriate clinical setting, is often difficult. While a chest roentgenogram compatible with interstitial lung disease (l/1 or greater profusion score) is generally accepted as excellent evidence of pneumoconiosis, the chest roentgenogram may be normal even in the presence of biopsy-proven pulmonary fibrosis. 1 As a

result, physicians have come to rely on tests of lung physiology such as pulmonary function tests (PFTs) to aid in the detection of restrictive lung function as an indication of the presence oflung fibrosis. The general pattern of abnormal physiology in restricted patients is well established, ie, decreased lung volumes, normal airflow, and reduced compliance with high maximal transpulmonary pressure. 2 Another PFT proposed as useful in the diagnosis of pneumoconiosis is the single breath diffusing capacity for carbon monoxide (Dco) or transfer factor of the lung. 3-s An abnormal Dco value may be the only resting physiologic factor to suggest clinical lung disease and in some studies the Dco value correlated with the severity of the histologic lesion in patients with asbestosis. 4 The reduction of the Dco value may precede abnormalities in the chest roentgenogram.3-5

*From the Departments of Internal Medicine, The University of Texas Health Center at Tyler, and Indiana University School of Medicine, Indianapolis. This study was supported in part by the Calvin H . English Endowment in Occupational Medicine, Indiana University. Dr. Garcia is a recipient of a Clinical Investigator Award from the National Heart, Lung and Blood Institute, K08 02312, and is the Calvin H. English Investigator in Occupational Lung Disease. Manuscript received April3; revision accepted December 12. Reprint requests: Dr. Garcia, 1001 W1st Tenth Street, OPW 425, Indianapolis 46202-2879

(Cheat 1990; 98:105-111) Dco =diffusing capacity; BAL = bronchoalveolar lavage; PFf =pulmonary function test; FVC =forced vital capacity; TLC =total lung_ capacity; P(A-a)01 =alveolar to arterial oxygen tension difference; FRC =functional residual capacity; GXT=graded exercise testing

CHEST I 98 I 1 I JULY. 1990

105

A reduced Dco value in an asbestos worker, in the absence of other known causes of diffusion capacity reduction, is considered by some investigators as presumptive evidence of asbestosis;3 however, the Dco value is affected by other factors such as cigarette smoking which may limit its apparent diagnostic value.6-ll Exercise testing with measurement of arterial oxygen saturation and bronchoalveolar lavage (BAL) are additional methods used to detect subclinical interstitial lung disease. Subjects with roentgenographic or histologic evidence of interstitial lung disease may have relatively normal pulmonary function at rest yet demonstrate significant abnormalities of gas exchange with exercise9 as well as exhibit BAL evidence of alveoli tis. 10 While the correlation between alveoli tis and reduced Dco values is unclear, 9 a relationship between Dco measurements and arterial desaturation with exercise has been suggested in certain disease populations such as selected individuals with chronic obstructive pulmonary disease (COPD) 11 and interstitial lung disease. 12 Because the significance of an isolated reduction in Dco values in patients with environmental dust exposure is uncertain, the current study examines the utility of exercise testing with gas exchange analysis and BAL in asbestos- and silicaexposed workers with a reduced Dco value as the only abnormality in their initial clinical evaluation and compares these results with those obtained in patients with comparable dust exposure and smoking history but normal Dco values. METHODS AND MATERIALS

Study Pvpulation and Clinical Evaluation The majority of patients whose cases are reported in this study were former employees of a glass bottle manufacturing plant who were referred to our ()(:cupational Medicine Clinic for medicolegal evaluation. Their occupational exposure hazards consisted of both asbestos and silica and the majority of the patients were exposed to both types of dusts. The protocol for the clinical evaluation of these patients was established in advance. All patients had a history and physical examination done by a pulmonary specialist, a complete set of PITs, chest roentgenograms showing four views of the thoracic cage (postero-anterior, lateral, and right and left obliques) that were classified according to ILOIUC 1980 standards by a B reader, without knowledge of the clinical findings, 13 and in most instances, BAL. Lavage was not performed in those patients with significant cardiac or pulmonary disease. A total ·of 286 patients were evaluated according to the protocol described above. From that overall patient population we retrospectively identified the subpopulation of patients that met the criteria for inclusion in this report. These criteria were as follows : (1) no roentgenographic evidence of pneumoconiosis; (2) FEV, percent predicted value greater than 70; (3) total lung capacity greater than 80 percent of the predicted value; and (4) no evidence of concomitant cardiovascular disease such as stroke, angina, or congestive heart failure. Patients who met each of these criteria could be separated into four groups (Table 1) according to their cigarette smoking history and whether or not they had a reduced Dco value (<75 percent of predicted value). Groups 1 and 2 consisted of nonsmokers

106

and groups 3 and 4 were composed of smokers. Groups 1 and 3 had normal Dco values (2!75 percent predicted), while groups 2 and 4 had reduced Dco values (<75 percent predicted). A number of patients in groups 1 through 4 underwent exercise testing as part of their clinical evaluation (n = 35).

PFTs Spirometric determinations were made using a dry rolling-seal spirometer (CPI model 220, Sensor Medic, Anaheim, Calif). Lung volumes were calculated from functional residual capacity (FRC) determined by plethysmography in a variable pressure, constantvolume plethysmograph (CPI 2000 Sensor Medic). Single-breath diffusing capacity for carbon monoxide (Dco) was performed using the technique of Ogilvie et al" and compared against normal values calculated from the equation of Miller et al" for all subjects. Predicted values for lung volumes were those of Boren et al•• for men and Goldman and Becklake 17 for women. Normal values for spirometric indexes were from Crapo et al.•• All PIT values were expressed as percent predicted values.

Exercise Testing Protocol An electronically braked cycle ergometer (model KEM-2, Mijnhardt, Odijk, Holland) delivered exercise work rate in increments of 10 W each minute, beginning with unloaded pedaling and increasing to the limit of tolerance.•• Patients maintained pedaling frequency above 50 cycles/min. Exercise data were recorded using an automated exercise testing system (model 2000, Medical Graphics Corporation, St. Paul, Minn) that converts breath-by-breath analog input to digital form. Expiratory Bow, gas concentrations, and 12-lead electrocardiographic (ECG) recordings were measured directly. Integration of the Bow signal from a pneumotachometer (model 3813, Hans-Rudolph, Kansas City, Mo) coupled with a pressure transducer (model 45-1-871 , Validyne Engineering Corporation, Northridge, Calif) provided quantitation of expired tidal volume. Routine volume calibration was accomplished with a 3-L calibrating syringe. Periodic checks using a 120-L Ttssot spirometer (Collins, Braintree, Mass) confirmed the accuracy of the system for measuring ventilation. Breathing frequency was computed in an automated fashion from cycle times of expired tidal breaths. All respiratory values were continually monitored on a breath-to-breath basis with the last 30 s of each one-minute stage averaged and expressed on a per minute basis. An infrared absorption analyzer (model CD-102, Datex Instrumentation OY, Helsinki, Finland) and a fuel cell analyzer (model S-3AII , Applied Electrochemistry, Sunnyvale, Calif) determined the fractional concentrations of carbon dioxide and oxygen, respectively, in expired gas. Gas analyzers were calibrated with a standard reference gas mixture (16 percent 0 1 , 6 percent COJ. Oxygen saturation values were obtained by either arterial blood gas measurement using a blood gas analyzer (model 178, Corning Instruments, Medfield, Mass) and a CO-oximeter (model 282, Instrumentation Laboratories), or by continuous monitoring with an ear oximeter (modei47201-A, Hewlett Packard Inc, Chelmsford, Mass, or Biox IIA Ohmeda). The ECG waveform and cardiac frequency were monitored throughout exercise using a 12-lead ECG (Q-2000, Quinton Instruments, Seattle, Wash).

BAL After the administration of local anesthesia to the upper airway, a fiberoptic bronchoscope was introduced into the lower respiratory tract and BAL was performed as previously described."' A total of 100 ml of sterile saline solution was injected in 20-ml aliquots into each of two lobes, the right middle lobe, and lingula. After instillation, the Buid was immediately aspirated using conventional wall suction and immediately after collection the Buid was filtered through gauze. A sample from each lobe was taken for cell Asbestos- and Silica-Exposed Workers (Garcia et al)

differential and cell number. The total volume of saline solutions instilled per patient was 200 ml, and the volume of lavage Huid recovered was typically 80 to 120 mi. A small aliquot of the BAL Huid was then used for cell differential counts and total white blood cell counts. Differential cell counts were determined from cytocentrifuge preparations (Cytospin II: Shandon-Southern Instruments, Sewickley, Pa) that were stained with a modified Wright-Giemsa stain (Harleco Diff-Quik, Dade Diagnostics, Aquada, PR). At least 300 cells were counted using morphologic criteria to enumerate the percent of macrophages, lymphocytes, neutrophils, and eosinophils present. Ciliated and squamous epithelial cells were not recorded. Total BAL white blood cell counts were obtained using a hemocytometer and total numbers of BAL macrophages, neutrophils, and eosinophils were obtained by multiplying the cell differential for each lobe by the total white blood cell count for that lobe. Total lung leukocytes were obtained by adding the results obtained for each lobe. Results were also expressed as the cells per milliliter of recovered BAL Huid. Statistical Analysis

All data are expressed as the mean± SEM . Comparisons between patient groups were calculated using the Mann-Whitney U test for BAL cell differentials and cell concentrations per milliliter of recovered BAL Huid. For all other outcome measures, either the pooled variance or separate variance t test was used pending the equality of group variances based on the F tests of sample variances. In addition, Spearmans rank order correlation coefficients were obtained for selected clinical variables. Significance was determined at p
lbtient Groups A total of 53 patients with Dco values less than 75 percent of predicted met the criteria for inclusion in this report. Thirteen of these patients were nonsmokers, group 2 (defined as not currently smoking and a total smoking history of less than five pack years with greater than five years since their discontinuation) and 40 were current or exsmokers, group 4 (>5 pack years). Control smoking (group 3, n = 35) and nonsmoking (group 1, n =37) subjects with normal diffusion values were identified in the original cohort. Individual Dco

125 -c 100 Q)

0

=o ~

a.

75

0 0 0 0

I

I

I

H

~ 0

0 0

0

I

8

50 25

I •

B

f "

0

1

2

Groups

4

3

FIGURE l. Individual values for diffusing capacity (Dco) (percent predicted) in each of the four patient groups. Group 1 = nonsmoker, normal Dco value; group 2 = nonsmoker, low Dco value; group 3 =smoker, normal Dco value; and group 4 = smoker, low Dco value. The mean values are found in Table l.

values for each of the groups are shown in Figure 1. The Dco values from groups I and 3 with normal diffusion were not significantly different from one another nor were groups 2 and 4 with reduced diffusion (Table I). All four groups were similar in terms of their age and occupational dust exposure (Table 1). Group 2 patients exhibited reductions in forced vital capacity (FVC) and total lung capacity (TLC) when compared with the other three groups (Table 1).

Exercise Testing Results of exercise testing are summarized in Table 2. Only a limited number of patients in groups 1 through 4 (n =35) underwent and completed the progressive, noninvasive exercise test. In addition, a limited number of patients in each of these groups had arterial blood gas measurements and calculation of their alveolar to arterial oxygen tension differences

Table !-Clinical and Pulmonary Function Dota in the Study Groupa*

No. of patients Age, yr Pack-years smoking Asbestos exposure, No. of patients Asbestos exposure, yr Silica exposure, No. of patients Silica exposure, yr FVC(%pred) FEV, (% pred) FEV,IFVC (%) Total lung capacity (% pred) Dco

Group 1 (NS;Normal Dco)

Group2 (NS;Low Dco)

Group3 (S;Normal Dco)

Group4 (S;Low Dco)

37 (34M,3F) 52.1 ± 1.7 0.5±0.2 33 21.4 ± 1.5 22 20.5± 1.7 98.2± 1.6 95.9± 1.9 78.4± 1.2 106.9± 1.5 89.3± 1.8

13 (12M,1F) 52.5±3.0 0.8±0.7 12 18.7±3.3 12 21.6±3.0 89.5±2.3t 88.8±2.7 79.9± 1.5 98.7±2.7t 64.6±2.5t

35 (34M,1F) 50.7± 1.9 35.4±4.5t:j: 31 20.1 ± 1.6 26 18.3±2.0 95.94± 1.9 91.1 ±2.2 75.8± 1.5 105.7± 1.8:1: 87.5± 1.9:1:

40 (38M,2F) 52.1±1.7 38.2±4.4t:j: 32 20.4 ± 1.7

*Dco =diffusing capacity; NS =nonsmoker; S =smoker; M=male; F=female; values represent (ii ± SEM). Normal tp
24

20.4 ± 1.7 95.5±2.2 87.6± 1.8t 75.2± 1.3:1: 106.9±2.l:j: 62.4± l.4t§ Dco~75

percent.

CHEST I 98 I 1 I JULY, 1990

107

Table 2-Ruting and E:rercise Gas Exchange lJGta in the Study Group

No. of patients V01 max (% pred) Resting Sao.(%) Exercise Sa01 (%) Resting VoNT Exercise VoNT Resting P(A-a)01 Exercise P(A-a)01

Group 1 (NS;Nonnal Dco)

Group2 (NS ;Low Dco)

5 79. 1 ::!:3.1 95.9::!:0.6 95.7::!:0.2 0.33::!:0.02 0. 15::!:0.01 4.2::!: 0 .9(3):1: 11.6::!:2.8(3)

6 74.4::!:4 .1 95.7::!:0.5 93.6::!:0.5 0 .31::!:0.04 0 .17::!:0.02 10.8::!: 5.4(3) 21.7::!: 7.0(3)

Group3 (S;Nonnal Dco) 11 86.3::!:5.3* 96.5::!:0.9 96.5::!:0.8 0.35::!:0.02 0.16::!:0.02 6.5::!:3.3(3) 8 .1 ::!:3.2(3)

Group4 (S ;LowDco) 13 69.8::!:5.0 95.1::!:0.9 94.9::!:0.7 0 .33::!:0.02 0. 18::!:0.02 10.2::!: 4.3(5) 15.5::!: 2 .6(5)

*p
(P[A-a]O~. In general, all of the exercise test results were normal in the patient groups tested. Two nonsmoking patients in group 2 with reduced Dco values and two smoking patients in group 4 with reduced Dco values had elevated alveolar to arterial 0 2 differences at rest and during exercise. Patients in group 4 had statistically significant reduction in the percent predicted oxygen consumption per minute (Vo2) capacity than their group 3 smoking counterparts with normal diffusion. None of the patients (n = 35) desaturated (>3 percent decline in percent 0~ during exercise.

BAL The volumes ofBAL fluid recovered per group were as follows: group 1 = 118 ± 13 ml; group 2, 121 ± 17 ml; group 3, Ill± 21 ml; and group 4, 108 ± 22 ml. Results of BAL are shown in Table 3. Both nonsmokers and smokers with reduced diffusion (groups 2 and 4, respectively) had greater numbers of total BAL cells in their lavage fluid and a greater number of macrophages, neutrophils, lymphocytes, and eosinophils than did their corresponding control groups with normal diffusion (groups 1 and 3). These differences were statistically significant for total numbers of BAL cells, alveolar macrophages, and lymphocytes in the nonsmokers. All comparisons remained unchanged

when the levels of total BAL cells and each BAL cell type were expressed as cells per milliliter ofBAL fluid recovered. For example, the total macrophages per milliliter of BAL recovered were 0.047 ± 0.005 for group 1, 0.083±0.02 for group 2, 0.107±0.02 for group 3, and 0.178±0.02 for group 4. Correlation Data

Correlation coefficients were calculated using Dco as the dependent variable and a number of independent variables for all four patient groups (n = 125) together and the nonsmoking (n =50) and smoking subgroups (n = 75) separately. The independent variables used included years of dust exposure, pack years, P(A-a)02 , spirometry and exercise indexes, and each of the total cell count variables in the lung lavage fluid . There was an inverse correlation between Dco and cigarette pack years (r=- .18, p=0.02), and a significant direct correlation between the Dco and either the FEV" FEF~75 , or the FRC. In nonsmokers the Dco correlated with the TLC (r = .28, p = .02). No significant correlations were found between the Dco values and the exercise variables tested either for the entire subject population (n = 44) or for any of the smoking and nonsmoking subgroups. In contrast there was significant and inverse correlation between Dco values and total BAL cells, total macrophages, total

Table 3- Total Numben of lnjlammatory and Immune Effector Cella in Bronchoalveolar Lavage (BAL) Fluid from the Study Groups BAL Results* Total BAL Cells Nonsmokers Group 1 (n = 34) Group 2 (n = 13) Smoke rs Group 3 (n = 29) Group 4 (n =37)

6 .09::!:0.8 1l.45::!:2.2t 13.0::!:2.7t 20.5::!:3.ot:l:§

Neutrophils

Lymphocytes

Eosinophils

Macrophage

0 . 14::!:0.02 0 .28::!:0.01

0 .42::!:0.07 0.86::!:0.35

0.03::!:0.01 0.11::!:0.1

5 .5::!:0.8 10.2::!:2.8t

0.41 ::!:0. 13 0.42::!:0.09t

0.53::!:0. 17 0.67::!:0.13§

0 .12::!:0.03 0.13::!:0.03

1l.9::!:2.4t 19.3::!:2.9t:j:§

*Total cells x 10"1200 ml of BAL instilled. tp
108

Asbestos· and Silica-Exposed Wori
Table 4-Correlation ofDco Value• with Bronchoalveolar Lavage (BAL) Cellular Parameten• vs Total BAL cells Groups 1-4 (combined) Nonsmokers Smokers

r=- .33 p= .001 r= - .38 p= .008 r= - .24 p= .03

vs Macs

vs PMNs

vs Lymphs

vs Eos

- .33 .001 -.37 .009 -.25 .03

- . 19 .03 - .25 .05 - . 12

- .20

.02 -.30 .03 -.15

-.16 .05 -.32

.10

.09

.02 - . 12 .10

*Macs = macrophages; PMNs =polymorphonuclear leukocytes; Lymphs= lymphocytes; and Eos, eosinophils.

neutrophils, lymphocytes, and total eosinophils in the combined groups as well as in the nonsmokers group (n = 47) (Table 4). Only total BAL cells and total macrophages were significantly and inversely correlated with Dco values in the smoking group (n = 66). Interestingly, when group 2 nonsmoking patients with reduced diffusion were analyzed separately, the greatest correlation between any BAL cell type and Dco value was for eosinophils (Dco vs percent eosinophil, r=- .53; Dco vs total eosinophil, r=- .64). DISCUSSION

The clinical diagnosis of pneumoconiosis usually depends upon a history of exposure and roentgenographic evidence of interstitial fibrosis or lung nodules. In the presence of these classic findings, the diagnosis is not difficult and pulmonary function studies usually confirm the presence of restrictive lung disease. When the chest roentgenogram is normal, the diagnosis of pneumoconiosis becomes more difficult, especially when the patient with occupational dust exposure has other potential causes for pulmonary symptoms such as cigarette smoking. In the absence of an abnormal chest roentgenogram, pulmonary function studies become much more crucial to the clinical diagnosis of pneumoconiosis. It is possible for pulmonary function studies to show the classic pattern of restrictive lung disease (reduced lung volumes, normal FEV 1 percent, reduced Dco) even in patients with normal chest roentgenograms; however, in our experience this combination of findings is rare and occurred in only four of the 286 patients referred to our clinic because of a history of exposure to either asbestos or silica during the past 24 months. The clinical combination we did find with some frequency, however, was the patient with a significant history of dust exposure, a normal chest roentgenogram, and a reduced Dco value as the only abnormality on PFT. Of the 286 patients we evaluated in our clinic during the time course of this study, there were 53 with normal chest roentgenograms and an isolated reduction of diffusion on their PFTs. One of the more important confounding variables in our patient population (as it is in many studies of pneumoconiosis) was cigarette smoking. Sixty-four

percent of our total patient population were current or exsmokers. Based on animal models of asbestosis, cigarette smoke exposure decreases particulate clearance and thus increases the fiber burden of the lung, potentially accelerating pathologic processes. In addition, cigarette smoking has been demonstrated to affect both exercise performance as well as the cellular constituents of the lower respiratory tract and is well associated with diminished Dco values in otherwise apparently normal individuals. We attempted to control for this variable to the extent permitted by using both smoking and nonsmoking groups. We investigated the significance of the isolated reduction in Dco values in patients with dust exposure in two ways. The first technique was to examine whether these patients exhibited abnormal gas exchange when stressed by graded exercise testing (GXT). Other investigators have found that a reduced Dco value correlates with abnormal gas exchange during exercise in selected patient populations, 11 • 12 although the relationship between Dco and gas exchange during GXT as reported in other studies is quite variable. 21 •22 Risk et al 12 performed GXT in 168 patients with a variety of interstitial lung disorders, including 18 patients with pneumoconiosis (asbestosis) and found a Dco value less than 70 percent of predicted value to be a good predictor of an increased P(A-a)02 during exercise. A much lower Dco value (<55 percent of predicted value) was required to predict desaturation in smoking patients with COPD. 11 In contrast, other exercise studies in patients with occupational dust exposure but without interstitial lung disease do not suggest that the Dco is helpful in predicting abnormal gas exchange. Agostoni et al 23 found only one of five asbestos workers with normal chest roentgenograms demonstrated ventilatory limitation during exercise and none of the five had exercise-induced arterial desaturation. 23 Sue et al 24 •25 attempted to correlate Dco values with gas exchange during exercise in shipyard workers exposed to asbestos. They found at least one abnormal test of gas exchange (VoNT, P[A-a]02 , or P[1-ET] C02) during exercise in 14 of 16 workers with a single-breath Dco value of <70 percent predicted; however, only 14 of 96 workers with abnormal exercise gas exchange had a low Dco value. 25 Thus, the diffusion capacity was a fairly specific but CHEST I 98 I 1 I JULY, 1990

109

not very sensitive predictor of abnonnal gas exchange during exercise.24 ·:zs Our attempts to use a reduced Dco value as a predictor of abnormal gas exchange during exercise were not successful. With the exception of mild reduction in Vo 2 percent predicted in group 4, our patients had essentially nonnal results of GXT studies and there was no correlation between Dco values and any of the exercise gas-exchange parameters. These results, at least for patients who were smokers, differ from those of Sue et al. z.t.:zs Although the pathophysiologic significance of a reduced Dco value is not completely understood, it has been postulated that it relates in some way to the ability of gas to transfer across the alveolaf"capillary membrane. Therefore, patients with a reduced Dco value would be expected to have an increased P(A-a)02 during exercise. In fact the P(A-a)02 was higher on average for the groups with low Dco values (groups 2 and 4) than the groups with normal Dco values (groups 1 and 3) both at rest and after exercise and four of the eight patients with low Dco values who were exercised with arterial blood gas detenninations had an increased P(A-a)0 2 • However, there was no correlation between the individual Dco values and the individual values for P(A-a)02 • The significance of these findings must be interpreted in light of two caveats. First, the number of patients studied with exercise is relatively small. Secondly, we performed only graded exercise with no steady-state measurements; thus, it may have been technically difficult to detect small changes in gas exchange. In addition to exercise testing, the significance of an isolated reduction in Dco value in our patients with dust exposure was evaluated by examining lower respiratory tract cellular content by BAL. Although a correlation of structure with function through actual tissue samples is ideal, it was obviously not possible in this group of patients. However, a number of studies have now shown that BAL is a satisfactory method for obtaining a representative cell population from the lung26 and that the presence of inflammatory cells in BAL predicts functional outcome in some of the interstitial lung diseases.27 Moreover, the current concept of the pathogenesis of most of the interstitial diseases, including the pneumoconioses, includes an important role for expansion of existing cellular constituents and the influx of inflammatory cells into the alveoli .211 •29 Recently, increased lower respiratory leukocytes in BAL from asbestos-exposed workers has been described .30 Furthennore, physiologic derangement (such as reduction in TLC and FVC) and roentgenographic abnormalities (worsening profusion score) significantly correlated with the percentage and total numbers of BAL neutrophils and eosinophils.30 In a separate but confirmatory report by Merrill and coworkers, 31 neutrophilic alveoli tis was found to pre110

diet functional deterioration among asbestos workers. Our data in occupationally exposed workers are consistent with numerous reports that human smokers have both an expanded alveolar macrophage pool that is greater than nonsmokers as well as significant increases in BAL neutrophils over their nonsmoking counterparts. 26 •211 •29 •32 Given this background we found that patients with reduced Dco values had greater numbers of total BAL cells, increased levels of immune effectors such as alveolar macrophages and lymphocytes, as well as increased numbers of total inflammatory cells, ie, neutrophil and eosinophils in their BAL fluid when compared with patients with normal diffusion values. These differences were statistically significant for total BAL cells and total BAL alveolar macrophages in the nonsmokers with impaired diffusion as well as for smokers with low Dco values. In contrast to the exercise data, there was significant correlation between Dco values and the BAL data. Thus, there was strong, albeit indirect, evidence that a low Dco value was associated with active alveolar inflammation. This suggests that a low Dco value and an abnormal BAL both may be related to the presence of alveolar inflammation and may be directly related to each other. This study does not provide sufficient information to answer its original question, ie, the significance of an isolated reduction in the single-breath Dco value in dust-exposed individuals with normal chest roentgenograms. The study is cross-sectional with no followup information and does not contain pathologic material to provide true structure-function correlation. The study does, however, provide limited insight as to the significance of this finding. The presence of inflammatory cells in the BAL fluid suggests the low Dco values may be a reflection of active inflammation in these patients that is independent of smoking habits. Although this may not be sufficient to establish a diagnosis of pneumoconiosis, it may indicate those patients who require closer follow-up. From the data presented, it would not appear that their follow-up requires the inclusion of noninvasive exercise testing. ACKNOWLEDGMENTS: The authors gratefully acknowledge the contribution and skills of Shanna Dodd, Mittie Marshall, Brigette Stoglin, Paula L. Garcia, Sara lriana, Sara Shepherd, Barhara Pruitt, Jean Cairns, and Jaynellen Wylie. REFERENCES Epler GR, McLoud TC, Gaensler EA , Mikus JP. Carrington CB . Normal chest roentgenograms in diffuse infiltrative lung disease. N Engl J Med 1978; 298:934-39 2 Fulmer JD , Crystal RG . Inte rstitial lung disease. In: Simmons DH , ed . Current pulmonary. Boston, Mass: Houghton Miffiin Co; 1979; 1:1-65 3 Murphy RL, Becklake MR, Brooks SM , et al. Diagnosis of non malignant diseases related to asbestos: official statement of the American Thoracic Socie ty. Am Rev Respir Dis 1986; 134:36368 Asbestos- and Silica-Exposed Workers (Garcia et al)

4 Murphy RLH, Sorenson K. Chest auscultation in the diagnosis of pulmonary asbestosis. J Occup Med 1973; 15:272 5 Weill H , Ziskind MM, Waggenspack C, et al. Lung function consequences of dust exposure in asbestos cement manufacturing plants. Arch Environ Health 1975; 30:88 6 Van Ganse WF, Ferris BG Jr, Cotes JE. Cigarette smoking and pulmonary diffusing capacity (transfer factor). Am Rev Respir Dis 1972; 105:30-41 7 Frans A, Stanescu DC, Veriter C, Clerbaux T, Brasseur L. Smoking and pulmonary diffusing capacity. Scand J Respir Dis 1975; 56:11?5-83 8 Rankin J, Gee JBL, Chosy LW. The influence of age and smoking on pulmonary diffusing capacity in healthy subjects. Med Thorac 1965; 22:366-74 9 Keogh BA, Lakatos E, Price D, Crystal RG . Importance of the lower respiratory tract in oxygen transfer. Am Rev Respir Dis 1984; 129:S76-S80 10 Keogh BA, Crystal RG. Pulmonary function testing in interstitial pulmonary disease: what does it tell us? Chest 1980; 78:856-SS 11 Owens G, Rogers R, Pennock B, Levin D. The diffusing capacity as a predictor of arterial oxygen desaturation during exercise in patients with chronic obstructive pulmonary disease. N Engl J Med 1984; 310:1218-21 12 Risk C, Epler G, Gaensler EA. Exercise alveol~arterial oxygen pressure differences in interstitial lung lung disease. Chest 1984; 85:69-74 13 International Labour Office: Guidelines for the use of ILO international classification of radiographs of pneumoconioses. Geneva, Switzerland: International Labour Office; 1980 14 Ogilvie CM, Forster RE, Blakemore WS, Morton JW. A standardized breath holding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Clio Invest 1957; 36:1-17 15 Miller A, Thornton J, Warshaw R, Anderson H, Teirstein A, Selikoff I. Single breath diffusing capacity in a representative sample of the population of Michigan, a large industrial state. Am Rev Respir Dis 1983; 127:27Q-77 16 Boren H, Kory R, Syner J. The Veterans Administration-Army cooperative study of pulmonary function , II: the lung volume and its subdivisions in normal men. Am J Med 1966; 41 :96-114 17 Goldman HI , Becklake MR. Respiratory function tests, normal values at median altitudes and prediction of normal results. Am Rev Tuberc Pulm Dis 1959; 79:457-67 18 Crapo R, Morris A, Gardner R. Reference values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 1981; 123:659-64 19 Carter R, Peavler M, ZinkgrafS, Williams J, Fields S. Predicting

20

21

22

23

24

25

26

27

28

29

30

31

32

maximal exercise ventilation in patients with chronic obstructive pulmonary disease. Chest 1987; 92:253-59 Garcia JGN, James HL, Zinkgraf S, Perlman MB, Keogh BA. Lower respiratory tract abnormalities in rheumatoid interstitial lung disease: potential role of neutrophils in lung injury. Am Rev Respir Dis 1987; 136:811-17 Burrows B, Strauss RH, Niden AH . Chronic obstructive lung disease, Ill: interrelationships of pulmonary function data. Am Rev Respir Dis 1965; 91 :861-68 Vandenbergh E, Billiet L, Van de Woestijne KP, Gyselen A. Relation between single breath diffusing capacity and arterial blood gases in chronic obstructive lung disease. Scand J Respir Dis 1968; 49:92-101 Agostoni P, Smith DD, Schoene RB , Robertson HR, Butler J. Evaluation of breathlessness in asbestos workers. Am Rev Respir Dis 1987; 135:812-16 Sue DY, Orne A, Hansen JE, Wasserman K. Lung function and exercise performance in smoking and nonsmoking asbestosexposed workers. Am Rev Respir Dis 1985; 132:612-16 Sue DY, Orne A, Hansen J, Wasserman K. Diffusing capacity for carbon monoxide as a predictor of gas exchange during exercise. N Eng) J Med 1987; 316:1301-06 Hunninghake GW, Gadek JE, Kawanami 0, Ferrans VJ, Crystal RG. Inflammatory and immune processes in the human lung in health and disease: evaluation by bronchoalveolar lavage. Am J Pathol1979; 97:149-206 Keogh BA, Hunninghake GW, Line BR, Crystal RG. The alveolitis of pulmonary sarcoidosis: evaluation of natural history and alveolitis dependent changes in lung function. Am Rev Respir Dis 1983; 128:256-65 Crystal RG, Gadek JE, Ferrans VJ, Fulmer JD, Line BR, Hunninghake GW. Interstitial lung disease: current topics of pathogenesis, staging, and therapy. Am J Med 1981; 70:542-68 Crystal RG, Bitterman PB, Rennard SI , Hance AJ, Keogh BA. Interstitial lung diseases of unknown cause: disorders characterized by chronic inflammation of the lower respiratory tract. N Eng) J Med 1984; 310:154-66, 235-44 Garcia JGN , Griffith DE, Cohen AB, Callahan KS. Alveolar macrophages from patients with asbestos exposure release increased levels of leukotriene B,. Am Rev Respir Dis 1989; 139:1494-1501 Merrill W, Cullen M, Care SB. Neutrophilic alveolitis and li.mctional deteriomtion among asbestos workers. Clio Res 1989; 37:479[abstract) Hunninghake GW, Crystal RG . Cigarette smoking and lung destruction: accumulation of neutrophils in the lungs of cigarette smokers. Am Rev Respir Dis 1983; 128:833-38

CHEST I 98 I 1 I JULY, 1990

111