Several studies have shown that two methods, quantitation of lung T lymphocytes by bronchoalveolar

Transcription

1 Thorax 1982;37 :19-25 Lung inflammation in saroidosis: omparison of serum angiotensin-onverting enzyme levels with bronhoalveolar lavage and gallium-67 sanning assessment of the T lymphoyte alveolitis I SHONBRGR, BR LIN, BA KOGH, GW HUNNINGHAK, RG RYSTAL From the Pulmonary Branh, National Heart, Lung, and Blood Institute, and Nulear Mediine Department, linial enter, National Institutes of Health, Bethesda, Maryland, USA ABSTRAT Serum angiotensin-onverting enzyme (A) is elevated in many patients with pulmonary saroidosis and has been proposed as a measure of disease ativity. The present study was designed to evaluate the possible relationship between serum A and diret measures of the intensity of the alveolitis of pulmonary saroidosis as measured by bronhoalveolar lavage and gallium-67 (67Ga) sans. To aomplish this, 64 measurements of serum A, lavage T lymphoytes, and lung uptake of 67Ga were performed in 41 patients with biopsy-proven saroidosis. levations of serum A were found on at least one oasion in 17 patients (41 %). However, serum A was found to be a poor preditor of the intensity of alveolitis in saroidosis as assessed by the quantitation of bronhoalveolar lavage ells that were T lymphoytes and by 67Ga sanning. levated serum A did not predit whih patients would have elevated proportions of lavage T lymphoytes, whih patients would demonstrate inreased pulmonary uptake of 67Ga, or whih patients would have high-intensity alveolitis as defined by a ombination of these riteria. These observations suggest that while serum A may be useful in diagnosing saroidosis, it does not reflet aurately the intensity of the alveolitis of the pulmonary omponent of this disease. Saroidosis is a hroni granulomatous disease of unknown aetiology that an affet nearly every organ in the body. Pulmonary involvement ours in more than 9% of patients; of this group, 2-25% suffer permanent loss of lung funtion and 5-1% die as a result.'-4 The earliest stage of pulmonary saroidosis is an alveolitis haraterised by a diffuse infiltration of the lung by large numbers of monoytes/marophages and ativated T lymphoytes.4-'2 urrent onepts of the pathogenesis of pulmonary saroidosis suggest that it is the ativated T lymphoytes that attrat monoytes to the lung to form building bloks for granulomata.13 In addition, the ativated lung T lymphoytes also stimulate lung B lymphoytes in a nonspeifi, polylonal fashion, thus ausing the Address for reprint requests: Dr l Shoenberger, Building 1, Room 6D-6, National Institutes of Health, Bethesda, Maryland 225, USA. hyperglobulinemia that is harateristi of the disease.14 Several studies have shown that two methods, quantitation of lung T lymphoytes by bronhoalveolar lavage and 67Ga sanning, an be losely orrelated with the alveolitis of pulmonary saroidosis The proportion of T lymphoytes found in bronhoalveolar lavage fluid orrelates with the intensity of the mononulear ell alveolitis observed morphologially and with the proportion of T lymphoytes that an be reovered diretly from lung biopsies. In addition, the intensity of uptake of 67Ga in the lung orrelates with the proportion of T lymphoytes reovered by bronhoalveolar lavage. Using these methods, patients with pulmonary saroidosis an be grouped into two ategories.4 16 Those with "high intensity alveolitis" are defined as those with lavage T lymphoytes of > 28 % together with a positive 67Ga san of the lung, and those with "low intensity alveolitis" are defined as those with 19 Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

2 2 lavage T lymphoytes of < 28 % or a negative 67Ga san or both. The utility of this lassifiation is illustrated by the prospetive evaluation of untreated patients with pulmonary saroidosis: of those with high intensity alveolitis, approximately 62 % will deteriorate in one or more lung funtion parameters over the subsequent six months. In ontrast, fewer than 8% of individuals with low intensity alveolitis will deteriorate over the same time period.16 In this ontext, the present study was designed to ompare the intensity of the alveolitis of saroidosis with measurements of serum A ativity, a test that is used in the diagnosis of saroidosis and has been suggested as a useful measure of the ativity of disease in these patients. To aomplish this, serum A was measured in 41 patients with biopsyproven pulmonary saroidosis and ompared with lavage and 67Ga estimates of the alveolitis made at the same time. Methods The study population onsisted of 41 individuals with biopsy-proven pulmonary saroidosis, with a mean age of years (all data presented as mean ± standard error of the mean). There were 16 men and 25 women; 19 were auasian and 22 were blak. Thirty-one (76%) were untreated at the time of entry into the study; the remainder were taking prednisone (mean dose 31-3 ± 2-8 mg per day). For the entire group, the mean duration of symptoms was 37- ± 5 9 months. Five patients were in radiographi stage (normal hest film); nine were in radiographi stage I (hilar adenopathy only); 11 were in stage II (hilar adenopathy plus pulmonary infiltrate), and 16 were in stage III (infiltrate only). Pulmonary funtion testing demonstrated mean values as follows: vital apaity 76-9 ± 2-7% of predited; total lung apaity 75.7 ± 2-7 % of predited; fored expiratory volume in one seond 819 ± 3-3% of predited; single breath diffusing apaity 86-2 ± 2-8% of predited (based on alveolar volume and haemoglobin); and fored expiratory volume in one seond/fored vital apaity of 14 ± 3-4% of predited.17 ANGIOTNSIN-ONVRTING NZYM Serum angiotensin-onverting enzyme was determined using the radiohemial assay developed by Ryan and o-workers18 and marketed by Ventrex Laboratories, Portland, Me. The substrate used was (3H)-hippuryl-glyyl-glyine, and enzyme ativity was measured as a funtion of the amount of (3H)- hippuri aid hydrolysed during a one hour inubation. nzyme ativity was expressed in arbitrary units (units = nanomoles hippuri aid generated/ Shoenberger, Line, Keogh, Hunninghake, rystal min-ml serum), with the normal range being units (mean 84 5 ± 2 SD). The normal range established by Ventrex has been validated by a number of laboratories and was onfirmed by our own laboratory using sera from 42 random healthy blood bank donors (21 male, 21 female). All patient data were determined using a frozen serum sample obtained by venepunture within 72 hours of the orresponding lavage and gallium studies. Standard referene sera as well as stored sera were used as ontrols in the assay. BRONHOALVOLAR LAVAG Bronhoalveolar lavage was arried out as desribed previously.5 8 The perentage of T lymphoytes in the lavage fluid was determined by rosette formation with neuraminidase-treated sheep red blood ells at GALLIUM SANNING Gallium sans were performed in a standard manner after the intravenous injetion of 5,ui/kg of gallium-67 itrate. The sans were evaluated using the semiquantitative "67Ga-index", as previously desribed,15 to minimise the effets of body morphology, imaging tehnique, and observer subjetivity on san interpretations. STAGING OF TH T LYMPHOYT ALVOLITIS Patients were grouped as having "high intensity alveolitis" or "low intensity alveolitis" using previously desribed riteria.4 High intensity alveolitis inludes individuals with lavage T lymphoytes >28 % of all reovered ells and a positive 67Ga lung san (67Ga-index >5 units). Low intensity alveolitis inludes individuals with lavage T lymphoytes < 28 % and/or a negative 67Ga san (67Gaindex < 5 units). Previous studies have demonstrated that these riteria orrelate well with morphologial assessment of the alveolitis and with prognosis of the lung disease-that is, stability or deterioration of lung funtion.4 16 In those individuals who were evaluated more than one, the lavage T lymphoyte and 67Ga-index parameters at the time of eah serum A value were used. Some individuals who had multiple serum A determinations were lassified as high intensity alveolitis on some oasions and low intensity alveolitis on others depending on the T lymphoyte and 67Gaindex values. DATA ANALYSIS Of the total group of 41 patients, 28 had single and 13 had multiple (range 2-5) serum A, lavage T lymphoyte, and 67Ga san determinations (total Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

3 Angiotensin-onverting enzyme in saroidosis 64 measurements of all three tests in 41 patients). Data on the total number of T lymphoytes reovered via bronhoalveolar lavage and the number of T lymphoytes/ml of lavage fluid were available in 24 patients, of whom 16 had single and eight had multiple (range 2-4) determinations of lavage T lymphoytes in onjuntion with simultaneous serum A and 67Ga sans (total 39 measurements in 24 patients). In those instanes where more than one measurement of eah test was made, eah was separated by three to 12 months. Several approahes were used for data analysis. I Serum A and lavage % T lymphoytes were ompared using the Pearson orrelation. This was done using eah patient only one (n = 41) and using all data from all patients (n = 64). 2 Serum A levels were ompared to lavage % T lymphoytes using the hi-square test. To do this, serum A levels were grouped as normal or elevated and the lavage % T lymphoyte data were grouped as < 28 % or > 28 %. This was done using eah patient one (n = 41) and using all data from all patients (n = 64). 3 Serum A was ompared with total lavage T lymphoytes reovered and T lymphoytes/ml of lavage fluid using the Pearson orrelation. In addition, these parameters were also ompared by separating the patients into two groups: those with normal (< 125 units) serum A, and those with elevated (> 125 units) serum A. For eah group, the mean and standard error of total T lymphoytes reovered and T lymphoytes/ml of lavage fluid were alulated. The respetive means for the two serum A groups were then ompared using the two-tailed Student's t test (that is, mean total T lymphoytes reovered in the normal serum A group versus mean total T lymphoytes reovered in the elevated serum A group; and mean T lymphoytes/ml of lavage fluid in the normal serum A group versus mean T lymphoytes/ml of lavage fluid in the elevated serum A group). In eah instane, these omparisons were done using eah patient only one (n = 24) and using all data from all patients (n = 39). 4 Serum A and 67Ga index values were ompared using the Pearson orrelation with eah patient used one (n = 41) and using all data from all patients (n = 64). 5 Serum A levels were ompared to the 67Ga indies using the hi-square test. To do this, serum A levels were grouped as normal or elevated and the 67Ga indies were grouped as < 5 index units or > 5 index units. This was done using eah patient one (n = 41) and using all data from all patients (n = 64). 6 Serum A was ompared to the intensity of the alveolitis by grouping the patients as having high intensity alveolitis or low intensity alveolitis and omparing the mean serum A levels of the two groups using the two-tailed Student's t test. This was done using eah patient one (n = 41) and using all data from all patients (n = 64). 7 Serum A levels were ompared to the intensity of alveolitis using the hi-square test. To do thus, serum A levels were grouped as normal or elevated and the patients were grouped as having high intensity alveolitis or low intensity alveolitis. This was done using eah patient one (n = 41) and using all data from all patients (n = 64). Results omparison of serum A levels with bronhoalveolar lavage demonstrated that serum A was not able to predit the proportion of lung inflammatory and immune effetor ells that were T lymphoytes (r = 2), p = -22, fig 1). This was true when eah patient was analysed one only (fig 1) and when all data from the 41 patients were onsidered (r = 17, data not shown). In addition, when patients were grouped as having < 28 % or >28% T lymphoytes in lavage (that level used to define high and low intensity alveolitis, respetively) serum A was unable to distinguish between the groups (p > -6 with eah patient evaluated one; p > -12 with all data from all patients) (table 1). Furthermore, no onventional linial or physiologial parameter predited whih patients would have % T lymphoytes of < 28 % or > 28 % (p > 1, all omparisons) (table 2). No signifiant orrelations were found between serum A and total T lymphoytes reovered, either with eah patient onsidered one (r = --16, p = 46) or using all data from all patients (r = --13, p = 42) (table 3). Similarly, no signifiant orrelations were found between serum A and T lymphoytes/ml of lavage fluid when eah patient was onsidered one (r = --12, p = 58) or using all data from all patients (r = - 14, p = 4) (table 3). In addition, when patients were grouped as having normal or elevated serum A, they showed no signifiant differenes in mean total T lymphoytes reovered (eah patient onsidered one, p > -6; all data from all patients, p > 5) or in T lymphoytes/ml of lavage fluid (eah patient onsidered one, p > 4; all data from all patients, p > 4) (table 3). omparison of serum A levels with 67Ga index also demonstrated that serum A was not able to predit aurately the uptake of 67Ga by the lung (r = -32, p = 4, fig 2). While this p value suggests signifiane, the r value of -32 for this 21 Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

4 22 n a) N ).) M a a) L- 3k _ 1 5 o o * Lavage % T Lymphoytes Fig 1 omparison between serum angiotensin-onverting enzyme (A) ativity and the alveolitis of saroidosis as measured by the proportion of lavage ells that are T lymphoytes. High normal value of serum A is 125 units. Normal T lymphoytes are < 8%; 28% is the ut-off level used in the lassifiation of high intensity alveolitis (see text for details). Patients whose 67Ga index was 5 index units are represented by open irles (). Patients whose 67Ga index was > 5 index units are represented by losed irles (). ah patient in the study is used one in this analysis (n = 41). There was no signifiant relationship between serum A and % T lymphoytes (r = -2, p > 21). Shoenberger, Line, Keogh, Hunninghake, rystal Table 1 omparison of serum angiotensin-onverting enzyme with lavage T lymphoytes, 67Ga sanning, and alveolitis intensity in patients with pulmonary saroidosis* Parameter Serum angiotensin-onverting enzyme (units) ah patient All data for all onsidered one patients onsidered only (n = 41) (n = 64) 6125 > 125 <125 > 125 % T lymphoytes 628% >28% '7Ga index -.S units >S units Alveolitis intensity High Low *Data presented as number of patients in eah group; hi-square analysis used. No signifiant differenes were found in any omparison (p > 1, all omparisons). omparison implies that serum A aounts for only 1% of the variation in gallium index. ven less orrelation was found when all data for all patients were analysed (r = -14, p = -26, data not shown). In addition, when patients were grouped as having 67Ga indies of < 5 or > 5 index units (that level used to define high and low intensity alveolitis, respetively), serum A was unable to distinguish between the groups (p > -22, all omparisons) (table 2). omparison of serum A levels with assessment of the alveolitis as measured by the ombined riteria of lavage % T lymphoytes and 67Ga index demonstrated that serum A was not able to predit the intensity of alveolitis whether eah patient was used only one (mean serum A ± SM for high intensity group 133 ± 13; mean serum A for low intensity group 11 ± 12, p = 23) (fig 3); Table 2 omparison of lavage % T lymphoytes, t7ga san, alveolitis intensity, and serum angiotensin-onverting enzyme with linial and physiologi parameters ofpatients with pulmonary saroidosis* Parameter linial Physiologialt Age Sex Rae Therapy: Duration oj V TL FV, FV, % DLO (yr) (M/F) (WIB) (YIN) symptoms (mo) % T lymphoytes 628% /15 13/1 7/16 4 ± 7* ± ± ± 3-2 >28% 35 ± 2-6 8/1 6/12 3/ ± ± 49 "Ga index.5 units 39 5 i 2-7 5/9 7/7 5/ ± ± ± ± 4-1 > 5 units 36-8 ± /16 12/15 5/ ± ± ± ± 3 9 Alveolitis intensity High 33-1 ± 2-9 7/7 4/1 1/ ± i ± ± ± 5-6 Low /18 15/12 9/ ± ± ± *7 ± ± 3-2 *All data presented as mean ± standard error of the mean; omparisons between means made under the two-tailed Student's t test; when disrete data are presented (ie, sex, rae, therapy), omparisons were made by hi-square test. tall data are given as % predited, see referene 13 for details; V: vital apaity; TL: total lung apaity; FVy: fored expiratory volume in one seond; FV,%: fored expiratory volume in one seond divided by fored vital apaity; DLO: diffusing apaity for arbon monoxide; DLO % predited is based on alveolar volume and haemoglobin. t(y/n); Y = on ortiosteroid therapy at the time of study; N = no therapy at the time of study. p < 5 between groups. Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

5 Angiotensin-onverting enzyme in saroidosis Table 3 omparison of serum angiotensin-onverting enzyme with lavage total T lymphoytes and T lymphoytes/ml in patients with pulmonary saroidosis A Regression analysis Parameter ah patient onsidered one only (n = 24) All data from all patients (n = 39) r p r p Total T lymphoytes vs serum A T lymphoytes/ml vs serum A B omparison of nmeans using Student's t test Parameter Serum angiotensin-onverting enzyme (units) Total T lymphoytes reovered ( x 1') T lymphoytes/ml ( X 12) ah patient All data from all onsidered one patients (n = 39) only (n = 24) <125 > 125 <S125 > ± ± ± 9 p > -6 p > ± -9 7 ± ± -2 p > -4 p > -4 or using all data from all patients (mean serum A for high-intensity group ; mean serum A for low intensity group 15 ± 8, p = 32, data not shown). In addition, when patients were grouped as having high intensity alveolitis or low intensity alveolitis using the ombined riteria of lavage % T lymphoytes and 67Ga index, serum A was unable to distinguish between groups (p > -23 with all data from all patients) (table 1). Similarly, no onventional linial or physiologial parameter, with the exeption of steroid therapy, predited whih patients would have high or low intensity alveolitis (p > 6, all omparisons) (table 2). There were signifiantly fewer patients in the high intensity group who were on therapy at the time of study (p < -5); this probably resulted from the fat that ortiosteroid therapy often onverts high intensity alveolitis to low intensity alveolitis.4 16 omparisons were also made between serum A levels, lavage % T lymphoytes, and 67Ga index on the one hand, and additional linial parameters inluding fever, sedimentation rate, haemoglobin, serum globulins liver funtion tests, night sweats, and radiographi stage on the other. Neither serum A, lavage % T lymphoytes, nor 67Ga index showed any signifiant orrelation with any of the linial parameters listed above (data not shown). Disussion Angiotensin-onverting enzyme is a dipeptidyl hydrolase that onverts angiotensin I to angiotensin If and inativates bradykinin. Sine the initial observation that erum angiotensin-onverting.t_ N w.q1 a) a) ) 4) l) s* o O * d I o 1 2 6"Ga Index (units) Fig 2 omparison between serum angiotensinonverting enzyme (A) ativity and the alveolitis of saroidosis as measured by the intensity of the 67Ga san. High normal value of serum A is 125 units. Normal 67Ga sans are those with 5 67Ga index units; > 5 index units is a neessary riterion for the lassifiation of high intensity alveolitis (see text for details). Patients whose lavage T lymphoytes were < 28 % are represented by open irles (). Patients whose lavage T lymphoytes were > 28 % are represented by losed irles (). ah patient in the study is used one in this analysis (n = 41). Although the relationship between serum A and the intensity of the gallium san has a p value of 4, an r value of -32 suggests this relationship aounts for only 1 % (r2 = -1) of the variability between these parameters. enzyme was elevated in pulmonary saroidosis,19 there has been a great deal of interest in the use of serum A as a measure of the ativity of this disease.2-26 However, diret omparison of serum A levels with the intensity of the alveolitis of pulmonary saroidosis as assessed by lung T lymphoytes and by 67Ga sanning has shown that although serum A levels are elevated in many patients with this disease, serum A does not reflet aurately disease ativity within the lung. The lak of orrelation between lung T lymphoytes and serum A in the present study was independent of the way in whih the T lymphoyte data were expressed-that is, perentage of total I Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

6 24 n. 6L) N ) (6. a) 6) 6) 3 25j_ 21_ 15 _ 1 _ 5 _ High Intensity Alveolitis. Low Intensity Alveolitis Fig 3 omparison betweent serum angiotensinonverting enzyme (A) ativity and the intensity of the T lymphoyte alveolitis ofpatients with pulmonary saroidosis as staged by bronhoalveolar lavage and 67Ga sanning. riteria for high intensity and low intensity alveolitis are listed in the text. ah patient in the study is used one in this analysis. There was no signifiant differene between the groups (p > -15). ells reovered by lavage that were T lymphoytes, total number of T lymphoytes in the lavage, or the number of T lymphoytes/ml of lavage fluid. Thus, in no ase ould a putative relationship between serum A and lung T lymphoytes explain more than 4 % of the data-that is, the largest r value was 2. ah of these approahes estimates a somewhat different aspet of the T lymphoyte alveolitis. Perent T lymphoytes expresses the relative emphasis of the lung's immune system toward T lymphoytes, while total T lymphoytes and T lymphoytes/ml of lavage express the total burden (or density per unit volume) of T lymphoytes within the alveolar strutures. Variability studies of lavage analysis have shown remarkable onsisteny in expressing ell data as a perentage of total lavage ells.5 8 In ontrast, there have been no Shoenberger, Line, Keogh, Hunninghake, rystal studies of the variability (from lobe to lobe or from day to day in a given patient) of expressing the ell data as the total number of eah ell type reovered or the number of eah ell type/ml of lavage fluid. In addition, although the average volumes of lavage fluid reovered among groups of patients tend to be similar, there is a wide satter in the volume reovered for eah individual. This variation is likely to be aused by lavage tehnique, airway geometry, and airway losure during aspiration, all of whih are quite independent of either diagnosis or disease ativity. Hene, while the number of T lymphoytes/ml gives a somewhat different view of the intensity of the alveolitis in saroidosis, there may be too muh variability in this approah to gauge aurately the total burden of T lymphoyte within the alveolar strutures. These fators may explain the onfliting results in the present study and others28 29 in terms of a orrelation between serum A and the numbers of T lymphoytes reovered per ml of lavage fluid. The failure of serum A to orrelate with the intensity of the alveolitis is saroidosis may be the result of a variety of fators. First, the soure of the elevated serum A is unlear. Studies have impliated alveolar marophages,3 31 irulating monoytes,32 and the epithelioid ells of granulomatous lymph nodes33-36 as possible soures of the inreased levels of serum A in some patients with saroidosis. In addition, two studies26 31 have demonstrated augmentation of A synthesis in tissue ulture in response to ortiosteroids, the very ompounds suspeted of lowering serum A levels in patients with saroidosis Seond, there is a well-doumented dihotomy between the peripheral and the pulmonary immune systems in saroidosis Suh divergene is also found in the literature speifially dealing with A: (1) A ativity in lung tissue/mg protein is normal in patients with elevated serum A33; (2) Serum A does not orrelate with levels of A produed by alveolar marophages from patients with pulmonary saroidosis maintained in vitro3; and (3) Serum A does not orrelate with levels of A in granulomatous lymph nodes from patients with saroidosis.36 Thus, while elevations of serum A are ommon in saroidosis, they are not universal, and do not orrelate diretly with the alveolitis of the disease. Measurements of serum A, although rapid, simple, and non-invasive, do not distinguish patients with high intensity from those with low intensity alveolitis, and thus represent a less preise method of staging these patients. Sine the alveolitis of saroidosis is not only the preursor of granuloma formation but also- represents the reversible om- Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.

7 Angiotensin-onverting enzyme in saroidosis ponent of the disease, it is more aurate to stage the alveolitis by lavage and by 67Ga sanning. Referenes l Mithell DN, Sadding JG. Saroidosis. Am Rev Respir Dis 1974;11: Mayok RL, Bertrand P, Morrison, Sott JH. Manifestations of saroidosis: analysis of 145 patients with a review of 9 series seleted from the literature. Am J Med 1963 ;35: Sones M, Israel H. ourse and prognosis of saroidosis. Am J Med 196;28: rystal RG, Roberts W, Hunninghake GW, Gadek J, Fulmer JD, Line BR. Pulmonary saroidosis: a disease haraterized and perpetuated by ativated lung T-lymphoytes. Ann Intern Med 1981 ;94: Weinberger S, Kelman JA, lson NA, Young R Jr, Reynolds HY, Fulmer JD, rystal RG. Bronhoalveolar lavage in interstitial lung disease. Ann Intern Med 1978; 89 : Hunninghake GW, Fulmer JD, Young R Jr, Gadek J, rystal RG. Loalization of the immune response in saroidosis. Am Rev Respir Dis 1979;12: Hunninghake GW, Fulmer JD, Young R Jr, rystal RG. omparison of lung and blood lymphoyte subpopulations in pulmonary saroidosis. In: Williams WJ, Davies BH, eds. Proeedings of the 8th International onferene on Saroidosis and Other Granulomatous Diseases. London: Alpha Omega, 198: Hunninghake GW, Gadek J, Kawanami, Ferrans VJ, rystal RG. Inflammatory and immune proesses in the human lung in health and disease: evaluation by bronhoalveolar lavage. Am J Pathol 1979;97: Hunninghake GW, Kawanami, Ferrans VJ, Young R Jr, Roberts W, rystal RG. haraterization of the inflammatory and immune effetor ells in the lung parenhyma of patients with interstitial lung disease. Am Rev Respir Dis 1981 ;123: rystal RG, Ferrans VJ, Gadek J, Fulmer JD, Line BR, Hunninghake GW. Interstitial lung disease: urrent onepts of pathogenesis, staging and therapy. Am J Med 1981 ;7: Daniele RP, Dauber JH, Rossman MD. Immunologi abnormalities in saroidosis. Ann Intern Med 198;92: Biserte G, hretien J, Voisin. Le lavage bronhoalveolaire hez l'homme. Paris: Inserm, Hunninghake GW, Gadek J, Young R Jr, Kawanami, Ferrans VJ, rystal RG. Maintenane of granuloma formation in pulmonary saroidosis by T-lymphoytes within the lung. N ngl J Med 198;32: Hunninghake GW, rystal RG. Mehanisms of hypergammaglobulinemia in pulmonary saroidosis. J lin Invest 1981 ;67: '5 Line BR, Hunninghake GW, Keogh BA, Jones A, Johnston GS, rystal RG. Gallium-67 sanning to stage the alveolitis of saroidosis: orrelation with linial studies, pulmonary funtion studies and bronhoalveolar lavage. Am Rev Respir Dis 1981;123: Keogh BA, Hunninghake GW, Line BR, Prie D, Young RG Jr, rystal RG. Alveolitis parameters as preditors of the natural history of pulmonary saroidosis. lin Res 1981 ;29:447A. 17 Fulmer JD, Roberts W, Von Gal R, rystal RG. Small airways in idiopathi pulmonary fibrosis: omparison of morphologi and physiologi observations. J lin Invest 1977;6: Ryan JW, hung A, Ammons, arlton ML. A simple 2* radioassay for angiotensin-onverting enzyme. Biohem J 1977;167: Lieberman J. levation of serum angiotensin-onverting enzyme (A) level in saroidosis. Am J Mied 1975;59: Ashutosh K, Keighley JFH. Diagnosti value of serum angiotensin onverting enzyme in lung disease. Thorax 1976;31 : Studdy P, Bird R, Geraint-James D, Sherlok S. Serum angiotensin-onverting enzyme (SA) in saroidosis and other granulomatous disorders. Lanet 1978;2: Turton GW, Grundy, Firth G, Mithell D, Rigden BR, Turner-Warwik M. Value of measuring serum angiotensin I onverting enzyme and serum lysozyme in the management of saroidosis. Thorax 1979 ;34 : Gupta RG, Oparil S, Szidon JP. linial signifiane of serum angiotensin onverting enzyme levels in saroidosis. JLab lin Med 1979;93: Lieberman J, Nosal A, Shlessner LA, Sastre-Foken A. Serum angiotensin onverting enzyme for diagnosis and therapeuti evaluation of saroidosis. Am Rev Respir Dis 1979;12: DeRemee RA, Rohrbah MS. Serum angiotensin onverting enzyme ativity in evaluating the linial ourse of saroidosis. Ann Intern Med 198;92: Baughman RP, Roberts RD, Ploy-Song-Sang Y. Relationship between serum antiotensin onverting enzyme and saroid ativity during steroid therapy. Am Rev Respir Dis 198;121:l IOA. 27 Reynolds HY, Fulmer JD, Kazmierowski JA, Roberts W, Frank MM, rystal RG. Analysis of bronho-alveolar lavage fluid from patients with idopathi pulmonary fibrosis and hroni hypersensitivity pneumonitis. J lin Invest 1977;59: Stanislas-Leguern G, Marsa J, Arnoux A, Leossier D. Serum angiotensin onverting enzyme and bronhoalveolar lavage in saroidosis. Lanet 1979;1: Rossman MD, ordillo M, Dauber JH, Daniele RP. orrelation between serum angiotensin-onverting enzyme and bronhoalveolar lymphoytes in saroidosis. hest 198;78:537S. 3 Hinman LM, Stevens A, Matthay RA, Reynolds HY, Gee JBL. Lysozyme and angiotensin onvertase ativity in human alveolar marophages: the effets of igarette smoking and pulmonary saroidosis. Am Rev Respir Dis 1978 ;117:67A. 31 Friedland J, Setton, Silverstein. Angiotensin onverting enzyme: Indution by steroids in rabbit alveolar marophages in ulture. Siene 1977;197: Friedland J, Setton, Silverstein. Indution of angiotensin onverting enzyme in human monoytes in ulture. Biohem Biophys Res ommun 1978;83: Silverstein, Friedland J, Lyons HA. levation of angiotensin onverting enzyme in granulomatous lymph nodes and serum in saroidosis: linial and possible pathogeni signifiane. Ann NY Aad S:i 1976;278: Silverstein, Friedland J, Akerman T. levation of granulomatous lymph node and serum lysozyme in saroidosis and orrelation with angiotensin-onverting enzyme. Am J /in Path 1977;68: Silverstein. Lokerman Z, Friedland J. Serum and lymph node ollagenase in saroidosis: omparison with angiotensin-onverting enzyme. Am J lin Path 1978 ;7: Silverstein. Friedland J, Lyons HA, Gourin A. Markedly elevated angiotensin onverting enzyme in lymph nodes ontaining non-nerotizing granulomas in saroidosis. Pro Natl Aad Si USA 1976;73: , 25 Thorax: first published as /thx on 1 January Downloaded from on 28 April 218 by guest. Proteted by opyright.