Bronchoalveolar Lavage in Alcoholic Liver Cirrhosis

Bronchoalveolar Lavage in Alcoholic Liver Cirrhosis* T-Lymphocyte Subsets and Immunoglobulin Concentrations Benoit Wallaert, M.D., EC.C .P.; Jean Frederic Colombel, M.D .; Lionel Prin, M.D .; Yves Sibille, M.D .; and Andre-Bernard Tonnel, M.D.

The purpose of this study was to determine the phenotype profiles of immune effector cells and the concentrations of immunoglobulins in the lower respiratory tract of nonsmoking patients with alcoholic liver cirrhosis (ALe). Nine nonsmoking patients with liver biopsy-proved ALC (grade B or C cirrhosis in Child's classification), free of clinical pulmonary symptoms, and with normal chest roentgenogram were included in the study. The control group included 12 healthy nonsmokers. Each patient had fiberoptic bronchoscopy with bronchoalveolar lavage (BAL). The number ofT cells and oflymphocyte subpopulations was determined by immunofluorescence studies using monoclonal antibodies that were specific for CD3, cm, and CD8 markers. Patients with ALC exhibited a dramatically increased percentage of CD8+ cells in BAL that induced a low CD4ICD8 ratio (0.96±O.15 vs 1.8±0.12 in healthy controls). Further characterization of lymphocyte subsets' dual immunofluorescence analysis demonstrated that most of the CD8+ alveolar lymphocytes had a phenotype of cytotoxic cells (CD8+ cnus-, 48 percent± 13 in ALC vs 10 percent±5

I n the last decade, several lines ofevidence suggested

that long-term alcohol intake and alcoholic liver cirrhosis (ALe) might be associated with an imbalance of the immune regulatory system .>' In this context, several immune abnormalities have been demonstrated in patients with ALe. Hypergammaglobulinemia, with elevations of IgA, IgM, and IgG, which is thought to be related to increased production, is a usual finding in patients with ALC.5.6 There is convincing evidence that development and expression of cell-mediated immunity are depressed in ALC. Patients with ALC have reduced skin test reactivity to ubiquitous antigens," Peripheral blood T cells, lymphocyte proliferation, suppressor cell function, and

*From the Departernent de Pneumologie, Hopital A. Calmelte, Lille (Drs. Wallaert and Tonnel), CJF INSERM 90-06 Instilut Pasteur, Lille (Drs. Wallaert and Tonnel); Service de Castroenterologie, Hopital Regional, Lille (Dr. Colombel); Laboratoire d'Immunologte, Faculte de Medecine, Lille (Dr. Pnn) , and Service de Pneumologie, Clinique UCL Mont-Godinne, Yvoir, France; and et Unite de Medecine Experimentale (ICP), Brussels, Belgium (Dr. Sibille). Supported by Fonds Special des Comites Departementaux contre la Tuberculose et les Maladies Respiratoires (87-C-6), by INSERM (Contrat Jeune Formation CJF 90-(6) and by Universtte de Lille II. Manuscript received January 11; revision accepted June 4. Reprint requests: Dr. Wallaert, Department Pneumolog/e, Hop/tal A. Calmette , 59037 Lille Cedes , France

468

in controls). ALC was associated with an appreciable alveolar-capillary "leak" as demonstrated by a significant increase in BAL 8uid albumin. In addition, the concentrations of immunoglobulins in BAL fluid were significantly greater in ALe than in controls . However, the relative (to albumin) coefficient of excretion of IgG, A, and M in and Os-macroglobulin BAL 8uid was DOt significantly different between controls and ALe. Our results indicate that increased proportions of CBS+ and especially of CDS + CDllb- cells are a common feature in the lower respiratory tract of nonsmoking patients with ALe. These changes may be of potential functional importance in the regulation of the local pulmonary immune response in ALe. (Cheat 1992; 101:468-73) ALC = aJc:oboIic liver cirrhosis; BAL= broocboalveolar lavage; F1TC 8uorescein isotbioc:yanate; HBSS Hanb balanced salt solution; IRMA = immunoradiometric assay; NK = natural killer; PBS = phosphate-buffered saline solution; ReE = relative coefficient of excretion

=

=

natural killer cell function are all diminished in ALC.7-18 However, wide variations were observed according to the studies suggesting that patient selection might be heterogenous since most of the investigations were reported without details about clinical status. Although extensive phenotypical and functional studies have allowed the characterization of a number of immune defects in their peripheral blood, no information is available about the local immune defense system in the lower respiratory tract of patients with ALC, It is known that the determination of lymphocyte phenotypes in peripheral blood does not provide information on the situation in specific organs. The fact that lymphocytes have the capacity to recirculate from blood to lymph and back to blood" supports the concept that lymphocyte recirculation might be important in the segregation oflymphocytes with distinct functions in the different lymphoid organs of the body. For example, comparison ofT cells obtained from bronchoalveolar lavage (BAL) to those circulating in peripheral blood has revealed large differences in various lung disorders.:lfi.22 Therefore, because patients with ALC experience an increased risk of lung infection, we initiated this study to BAL in Alcoholic LiYllrCln1losls(Wallaert et til)

Table l-Charocteriatica ofthe 12 Nommoking Patients with Alcoholic Liver Cirrhosis·

Patient

Age, yr

Child's Classification

Albumin, gll00 Iii) (3-5.4)

1 2 3 4 5 6 7 8 9

42 61 54 55 54 58 49 36 44

B B B B B C C C C

2.9 2.9 3.4 3.4 3.1 3.5 4.5 3.6 3.9

19A, gtlOOml (0.9-4.5)

IgG , gluXi ml (0.9-2)

IgM , iVl00 ml (0.6-2.5)

Bilirubin, gllOOml (0.2-1.2)

scar,

5.6 5 .8 3.7 5.3 4.2 3.5 3.7 5.4 2.1

2.6 2.6 1.5 1.9 1.2 0.9 1.5 1.4 1

3.5 6.3 2.4 8.7 2.1 1.1 2.2

0.5 2.5 1.5 2.5 1.7 0.9 0.8 0.7 1.4

140 146 129 146 161 148 167

1.2 1.6

lU/ml (7-40)

220 136

Alkaline Phophatase, lUll (50-180) 144 133 77 133 303

126 150 120 184

·Numbers in parentheses are normal values.

characterize lymphocyte phenotype profiles in the BAL and peripheral blood from nonsmoking patients with ALC using cell surface marker phenotyping by monoclonal antibodies. In addition, using highly sensitive techniques to analyze the secretions of plasma proteins in external body 8uids,23.24 we measured the concentrations of albumin, IgG, IgA, IgM and ~­ macroglobulin in serum and in BAL fluid, METHODS

lbtients Nine nonsmoking patients with biopsy-proved alcoholic cirrhosis (grade B or C cirrhosis in Child's classification) were included in the study (Thble 1). They were six men and three women, with a mean age of 51 ±3 years. The followmg criteria were used for inclusion in the study: (1) presence of at least one major sign of cirrhosis (jaundice, ascites, or gastrointestinal hemorrhage); (2) absence of infection; and (3) absence of surgery, antibiotherapy, or corticosteroid therapy for one month before hospital admission. None of them had signs and symptoms of acute alcoholic intoxication. Patients with ALC had a greater than 8O-g daily intake of alcohol for more than ten years. All the patients had not ingested alcohol for at least one week before these studies were performed and blood alcohol levels were zero at the time of the study. All were nonsmokers, had normal chest roentgenograms, and did not present evidence of lung infection three months before evaluation. None had signs of severe malnutrition. All gave informed consent for the procedure. Results were compared with those obtained from 12 age-matched healthy nonsmokers. Bronchoalveolar Lavage Bronchoalveolar lavage was performed after premedication with atropine under local anesthesia with lidocaine (lignocaine) using a wedged fiberoptic bronchoscope (model BF-B3; Olympus Corp of America, New Hyde Park, NY) and 250 ml of sterile saline solution was applied in five 5O-ml a1iquots with immediate gentle vacuum aspiration after each aliquot, as previously described." The aspirated Ruid was collected into siliconized jugs and immediately transported on ice to the laboratory. BAL was filtered through several layers of sterile surgical gauze and the cells were separated from the Ruid by low-speed centrifugation (10 min , 800 g). After three washings, the cells were resuspended in 10 ml of Hanks balanced salt solution (HBSS) and evaluated for their total number using a hemocytometer. A differential cell count (made from a total of 300 cells) was accomplished by morphologic criteria in cytocentrifuged smears stained with Wright-Giemsa. The viability of BAL cells as assessed by trypan blue dye exclusion was consistently greater than 90 percent in each experiment.

Identification of Lung and Blood lJ,pnphocyte Subpopulations The BAL cell pellet was resuspended and the cells were washed twice in HBSS and finally resuspended at 5 .10"/ml in RPMI 1640 (Eurobio) . BAL T lymphocytes or peripheral T cells were identified by indirect immunoRuorescence. BrieRy, 200 JLI of cell suspension was first incubated with 5 mgIL of unlabeled antibody (anti-CD3 and anti-eD4 from Ortho Pharmaceutical, Raritan NJ; anti = CD8 and anti=CDllb from Becton Dickinson, Meylan, France) for 30 min at 4°C. After three washings with HBSS, cells were incubated with Ruorescein isothiocyanate (FITC) labeled (Fab'2) fragment of antimouse IgG antibody (Dynatech Paris, France) for 30 min at 4°C and were washed twice with HBSS . The cells were resuspended in RPMII640and examined using a Ruorescence microscope equipped with phase contrast optics (Olympus BHS , Tokyo, Japan). The percentage of Ruorescein-Iabeled lymphocytes was calculated after counting a minimum of 200 lymphocytes per slide. Cell subpopulations were additionally characterized using the same method by their reactivity with anti-CD4 (Ortho Pharmaceutical, Raritan, NJ) and anti-eOS (Becton Dickinson, Meylan , France). CDllb has been proposed to distinguish among CD8' lymphocytes with suppressor activities (CD8' , CD ns-) from those with cytotoxic activities (CD8' , CDllb·)...... To simultaneously determine the coexpression ofCD8 and CDllb markers, a double staining technique was performed by using monoclonal antibodies conjugated to either FITC (COS) or to the RDl phycoerythrin derivative (CDllb) obtained from Becton Dickinson. BAL cells were incubated with 5 JLI of labeled antibodies for 30 min at 4°C washed three times with phosphate-buffered saline solution (PBS) at 4°C. The number of positive cells was determined by using Ruorescence microscopy. The surface phenotype of peripheral blood lymphocytes was determined using the procedure described above on blood mononuclear cells isolated by FicolVHypaque sedimentation. All results were expressed as the percentage of total lymphocytes. fuJtein Assay An immunoradiometric assay (IRMA) was used for measurement of the proteins in BAL. This assay previously described in detail , provides a sensitivity in the range of ng-ml" I and was performed on nonconcentrated BAL Ruid. ID ... BAL samples were diluted in 20 percent goat serum in PBS, at a pH of 7.4 . Serum levels of the different proteins were determined by immunonephelometry. Results are expressed in absolute concentrations in BAL (milligrams per liter) and in terms of the relative coefficient of excretion (RCE), which expresses the secretion rate of each protein relative to that of the entirely plasma-derived albumin as follows: RCE = [(protein BAUprotein serum)/(a1bumin BAUalbumin

serumr] CHEST I 101 121 FEBRUARY, 1992

489

Table 2-Diatribution ofBronchoalveolar Cella* Healthy Controls (n= 12) Recovered fluid, ml Total cell number, 100/mi Alveolar macrophages, % Lymphocytes, %

12

IS.4~2 .3

124±6 21±4.2

91.1~2 .2

84.3~1.St

9.9~1.5

16.1±2.1t 0.S±0.2

138~

Neutrophils, % Eosinophils, %

Patients with AlcoholicLiver Cirrhosis (n =9)

0 .5~0.2

o

0 .06~0 .06

*Results are expressed as mean ± SEM . tSigni6cantly different from healthy controls (p<0.05).

Statistical Analysis When applicable, data were expressed as mean ± standard error of the mean (SEM). Because most of the data were nonparametric, results were compared using the Mann-Whitney U test, Wilcoxon signed rank test, and Spearman rank coefficient. Only p values less than 0.05 were considered signi6cant. RESULTS

Bronchoscopic Findings

Fiberoptic bronchoscopy showed normal airways in all subjects without evidence of bronchial infection or inflammatory disease. There was no significant difference in percentage return of lavage 8uid between patients and controls. Number and Types of Bronchoalveolar Cells

Results of BAL total and differential cell count are summarized in Table 2. Total recovery cell yield did not significantly differ among patients and controls. Evaluation of the BAL cell differential of the patients with ALe demonstrated a slight but significant increased proportion of lymphocytes. Alveolar macrophage proportions were thus altered accordingly, and were lower than normal in the ALe group. The mean proportion of neutrophils and of eosinophils was not different between patients with ALC and controls. Characterization of Lung and Blood Lymphocyte Subpopulations

Identification of lung lymphocyte subpopulations showed that BAL lymphocytosis was predominantly

composed of T lymphocytes (Table 3). Analysis of Tcell subsets pointed out that most of the T cells in patients with ALe expressed the CDB suppressor/ cytotoxic phenotype, with a marked reversal of the CD4ICDB ratio (O.96±O.15 vs l.B±O.12 in healthy nonsmokers, p
The purpose of this study was to gain some insight

Table 3-Surface Phenotype ofLymphocytes Recoveredfrom Bronchoalveolar lAvage (BAL) andfrom Blood ofNonsmoking lbtient8 with Alcoholic Uver CirrhoBia and Healthy Nonsmokenr*

BAL Alcoholicliver cirrhosis Controls Blood Alcoholicliver cirrhosis Controls

C03· , %

C04· , %

COS·, %

COBb· , %

COS·COBb· , %

COS· COllb- , %

66~6

3O±15 46±12

51±16t 35± 1.4

4±4* 24±7

2±3.4* 25±7

48± 13* 10±5

43~15

26±10 29±S.9

IS±ll 22±7

22±7

64±4 64~20

67.1±9

42±10

16~S

13±10 10±6

*Results are expressed as mean ± SEM . tSigni6cantly different from controls (p<0.05). *Signi6cantly different from controls (P
470

BALin AJcohoIic LiYerCirThosis (Wallaert fit III)

Table 4-Protein Concentrations in the Bronchoalceolar LAvage (BAL) Fluid of Controls and Patients with Alcoholic Liver Cirrhosis· Groups

n

Alhumin

IgG

I~

IgM

C o nt rols Alcoholi c liver cirrhosis

13 9

26 (5) 44 .2 (5 .8)t

3.7 (0.5) 7.9 (2 .5)t

0.056 (0.08) 2.13 (0.82)t

0.017 (0. 004) 0.60 (0 .32)t

0 .04 (0.0 15) 0.06 (0 .03)

·Concentrations a re milligrams per liter of unconcentrated BAL fluid . Results a re expressed as mean (SE M). a , M = alpha-macroglobulin. t p< 0.5 compared with control value s. tp
into the immune system of the lower respiratory tract in ALC through evaluation of BAL T lymphocytes with monoclonal antibodies and of BAL fluid immunoglobulin. 1;' eliminate confounding factors that could have independent effects, we studied nonsmoking patients. Our results demonstrate that patients with ALC have a selective increase in the number ofCD8 + T lymphocytes in the lower respiratory tract which is not associated with similar abnormalities in the blood . The mechanisms responsible for the shift in the CD4ICD8 halance in the lower respiratory tract in ALC is unclear. One cannot exclude that a subclinical infection with viruses or Pneumocystis carinii might have resulted in the abnormalities observed herein. However, our patients with ALC were asymptomatic, had clear chest roentgenograms, and cytologic examination of BAL cells and bacteriologic analysis of BAL fluid did not demonstrate evidence of pulmonary infection. Our group and others'":" previously demonstrated that nonalcoholic patients with primary biliary cirrhosis exhibited an increased percentage of CD4 + T cells in BAL. These findings suggest that expansion of CD8 + T cells in ALC was not caused by either the underlying cirrhotic process or its biologic consequences since the patients were similar with regard to Child's classification and biologic associated serum abnormalities. In addition, it is unlikely that

the abnormalities found in patients with ALC are due to the effect of malnutrition since our patients did not exhibit signs of severe malnutrition. Moreover, in a previous study of patients with Crohns disease and who demonstrated various degrees of malnutrition, BAL studies showed an expansion of CD4 + T lyrnphocytes." Our results are in agreement with those obtained by immunohistologic study of liver biopsy specimens. Si and coworkers" demonstrated that the T lymphocytes accumulating in hepatic tissues of patients with alcoholic liver disease consisted mostly of CD8 + suppressor/cytotoxic cells. However, one cannot exclude that this local relative hepatic increment in suppressor/cytotoxic T cells might be something unique for hepatic inflammatory processes in general which may activate CD8 + cells and attract them to sites within liver lesions. Indeed, CD8+ T cell increment has been reported also in chronic active hepatitis and primary biliary cirrhosis." CD8 is a surface protein bound on cytotoxic and suppressor T cells. A local excess of suppressor/ cytotoxic cells might result either from nonspecific activation of the immune system by mononuclear cells or from activation by cytokins produced within the lung or may reflect an increase in natural killer (NK) cells that may bear more than one phenotype. 35 Using

1.2 -r----------------------, I-fI)

z~ WW -I20

1.0

W 0-' C U lEI

0.8

a:

II. II.Q,

II.

~O

~::)w lEI

u

0.6

~a:

0-I-Z

--2

0.4

wti

>a: -u 0.2 1-)( ~w wli. a: O

0.0 IgG

IgA

IgM

alpha 2 M

FIGURE l. Relative (to alhumin) coefficient of excretion (ReE) of immunoglobulins in the bronchoalveolar lavage (BAL) of healthy nonsmokers and of nonsmoking patients with alcoholic liver cirrhosis (ALC). Results are expressed as mean ± SEM.

CHEST I 101 I 2 I FEBRUARY, 1992

471

a dual immunofluorescence analysis, we were able to distinguish between the suppressor and cytotoxic CD8+ subsets. We demonstrated that most of the CD8+ alveolar lymphocytes had a phenotype of cytotoxic (CD8+ CDllb-) T cells. The fact that peripheral blood T-Iymphocyte studies do not demonstrate imbalance in T-cell subpopulations suggests that accumulation of CD8+ CDllb- T cells in the lower respiratory tract was not due to a cellular redistribution from the peripheral blood to the lung. However, we cannot rule out the possibility that we are dealing with specific cytotoxic cells. Although the CD8+ nsphenotype is consistent with the surface pattern of cytotoxic T lymphocytes, it is clear that there is not complete concordance between surface phenotypes and functional activities. Specific functional tests are needed to determine whether lung lymphocytes in ALC serve as functionally active cytotoxic cells. Evaluation of BAL fluid proteins demonstrated an appreciable alveolar-capillary "leak" as evidenced by a significant increase in lavage fluid albumin. Because of the efficiency of the epithelium barrier in the lower respiratory tract of healthy subjects, concentrations of plasma proteins of medium molecular weight are much lower in the epithelial lining fluid than in the plasma. In addition, because of the molecular-size affected seepage of plasma proteins across the epithelium barrier, IgM (900,000 daltons) concentrations in BAL fluid are usually very low. The present study demonstrated that ALC was associated with an increased movement of molecules such as albumin and immunoglobulins from the blood to the epithelial surface of the lower respiratory tract. Increased albumin and immunoglobulin concentrations in BAL are usually due to a pronounced leakage of the alveolar capillary barrier altered by inRammatory processes." However, although the RCE of IgG, 19A, and IgM was normal in ALC, one cannot exclude that a local immunoglobulin secretion in the lower respiratory tract might account, at least in part, for the increased concentration of BAL immunoglobulins. In this context, it is of importance that the BAL Cl2 M concentration in ALe was not significantly different from that of controls. The fact that the molecular weight of IgM and ~M is similar suggests that IgM might be locally secreted. Thus, an increased production of immunoglobulins in the lung of patients with ALC might contribute to the increased concentrations of immunoglobulins both in BAL and in serum, as was reported in interstitial lung disease." It is of particular interest to point out that similar observations have been found in the intestinal mucosa of alcoholics. Bjarnason and coworkers" clearly demonstrated that alcohol abuse was associated with significant changes in intestinal permeability leading to the concept of the leaky gut ofalcoholism . Moreover, 472

although densities of immunoglobulin-producing cells in the jejunal lamina propria of alcoholic patients were normal, these patients exhibited an increased permeability to plasma proteins such as albumin and irnmunoglobulins." In conclusion, our results demonstrated that ALe is associated with profound imbalance of the immune respiratory T cells. Although mechanisms responsible for local immune disturbances are unknown, expansion of CD8+ l lb : cytotoxic cells might be of crucial importance in the defense mechanisms of the lower respiratory tract in patients with ALe. ACKNOWLEDGMENT: Nadine Duhautois and Veronique Lequeux provided technical assistance.

REFERENCES 1 Heinemann HO. Alcohol and the lung. Am] Med 1977; 63:8185

2 Zetterman RK, Sorrell MF. Immunologic aspects of alcoholic livers disease . Gastroenterology 1981; 81:616-624 3 MacGregor RR: Alcohol and immune defense . JAMA 1986; 1474:9 4 Bomalaski IS, Phair ]P. Alcohol, imunosuppression, and the lung. Arch Intern Med 1982; 142:2073-74 5 Feizi T. Immunoglobulins in chronic liver disease . Gut 1968; 9:193-98 6 Wilson ID , Onstad G, Williams RC. Serum immunoglobulin concentrations in patients with alcoholic liver disease . Gastroenterology 1969; 57:59 7 Berenyi MR, Strauss B, Cruz D. In vitro and in vivo studies of cellular immunity in alcoholic cirrhosis. Am ] Dig Dis 1974; 19:199-205 8 Thomas HC. T cells subsets in patients with acute and chronic HBV infection, primary biliary cirrhosis and alcohol induced liver disease. Int] Immunopharmacoll981; 3:301-05 9 Alexander G]M, Nouri Aria KT, Eddleston ALWF, Williams R. Contrasting relations between suppressor-cell function and suppressor-cell number in chronic liver disease . Lancet 1983; 2:1289-93 10 Perrin D, Bignon ]D, Beaujard E, Cheneau ML. Populations de lymphocytes T circulants chez les patients atteints de cirrhose alcoolique. Gastroenterol Clin BioI 1984; 8:907-10 11 Pelletier G, Segond P, Attali P, Briantais M], Etienne]P. Etude phenotype des lymphocytes T sanguins au cours des hepathopathies alcooliques. Gastroenterol Clin BioI 1984; 8:911-14 12 Keever UMC, Mahony CO, Whelan CA, Weir DG, Feighery C. Helper and suppressor T lymphocyte function in severe alcoholic liver disease . Clin Exp Immunoll985; 60:39-48 13 Pirroni S, Tosato F, Rossi P, Fossaluzza V, Tonutti E , Sala PG. T cell subsets in peripheral blood and ascitic 8uid of patients with alcoholic liver cirrhosis. Lancet 1983; 2:518-21 14 ]ovanic R, Worner T, Lieber CS, Paronetto F. Lymphocyte subpopulations in patients with alcoholic liver disease. Dig Dis Sci 1986; 31:125-30 15 Kalrumu S, Leevy CM. Lymphocyte cytotoxicity in alcoholic hepatitis . Gastroenterology 1977; 72:594-97 16 Kawaniski H, Tavassolie Rp, Mac Dermott Rp, Sheagran ]N. Impaired concanavalin A inducible suppressor T cell activity in active alcoholic liver disease. Gastroenterology 1981; 80:510-17 17 Gilhus NE, Matre R. In vitro effect of ethanol on subpopulations of human blood mononuclear cells. Int Arch Allergy Appl Immunol 1982; 68:382-86 18 Ikeda T, Daiguji Y, Hasumara Y, 'Thkeuchi ]. In vitro effect of prednisolone on peripheral blood suppressor T cell activity in BAlin AlcoholIc Uver Clntlosis (Wallaert et aI)

19

20 21

22

23

24

25

26

27

28

29

patients with alcoholic hepatitis. Clin Immunol Immunopathol 1989; 53:225-32 Deviere J, Denys C, Schandene L, Romasco F, Adler M, Wybran J, et al. Decreased proliferative activity associated with activation markers in patients with alcoholic liver.cirrhosis. Clin Exp Immunoll988; 72:377-82 Gowans JL , Knight EJ. The route of recirculation oflymphocytes in the rat . Proc R Soc Lond B BioI Sci 1964; 159:257-82 Hunningghake GadekJE , YoungRD , Kawanami 0, Ferrans VJ, Crystal RG. Maintenance of granuloma formation in pulmonary sarcoidosis by T lymphocytes within the lung. N Engl J Med 1989; 302:594-98 Leathermann MD, Michael AF, Schwartz BA, Hoidal JR. Lung T cells in hypersensitivity pneumonitis. Ann Intern Med 1984; 100:390-94 Wallaert B, PrinL, Hatron PY, Tonnel AB, VoisinC . Lymphocyte subpopulations in bronchoalveolar lavage in SjOgren's syndrome: evidence for an expansion of cytotoxic/suppressor subset in patients with alveolar neutrophilia. Chest 1987; 92;1025-31 Delacroix DL, Hodgson HJ, McPherson A, Dive C, Vaerman JP. Selecti~e transport of polymeric immunoglobulin A in bile: quantitative relationships of monomeric and polymeric immunoglobulin A, immunoglobulin M and other proteins in serum, bile and saliva. J Clin Invest 1982; 70:230-41 Delacroix DL, Marchandise FX, Francis C, Sibille Y. Alpha-2macroglobulin, monomeric and polymeric immunoglobulin A, and immunoglobulin M in bronchoalveolar lavage. Am Rev Respir Dis 1985; 132:829-35 Wallaert B, Aerts C, Bart F, Hatron PY, Dracon M, Tonnel AB, et al. Alveolar macrophage dysfunction in systemic lupus erythematosus. Am Rev Respir Dis 1987; 136:293-97 Reinherz EL, Schlossman SF. Regulation of the immune response inducer and suppressor T-Iymphocyte subsets in human beings. N Engl J Med 1980; 303:370-74 Ledbetter]A, Evans RL, Lipinski M, Cunningham-Bundles C, Good RA, Herzenberg LA. Evolutionary conservation of surface molecules that distinguish T lymphocyte helper/inducer and cytotoxic/suppressor subpopulation in mouse and man . J Exp Med 1981; 153:310-16 Clement LT, Dagg MK, Lauday A. Characterization of human

ow

lymphocyte subpopulations: alloreactive cytotoxic T-Iymphocyte precursor and effector cells are phenotypically distinct from Leu 2+ suppressor cells. J Clin Immunoll984; 4:395-401 30 Gatenby PA, Kansas GS , Xian SY, Evans RL, Engleman EG . Dissection. of immunoregulatory subpopulations of T lymphocytes within the helper and suppressor sublineages in man . J Immunoll982; 129:1997-2001 31 Wallaert B, Bonniere PH, Prin L, Cortot A, Tonnel AB, \hisin C. Primary biliary cirrhosis: subclinical inflammatory alveolitis in patients with normal chest roentgenograms. Chest 1986:

90:842-48 32 Spiteri MA, Johnson M, Epstein 0, Sherlock S, Clarke S\v, Poulter LW Immunological features oflung lavage cells patients with primary biliary cirrhosis may re8ect those seen in pulmonary sarcoidosis. Gut 1990; 31:208-12 33 Bonniere PH, Wallaert B, Cortot A, Marchandise X, Riou Y, Tonnel AB, et al. Latent pulmonary involvement in Crohns disease: biological , functional bronchoalveolar lavage and radiographic studies. Gut 1986; 27:919-25 34 Si L, Whietesido TL, Schade RR, Thiel DY. Lymphocyte subsets studied with monoclonal antibodies in liver tissues of patients with alcoholic liver disease. Alcohol Clin Exp Res 1983; 7:431-35 35 Husby G , Blomhoff JP, Elgjo K, William s RC Jr. Immunohistochemical characterization of hepatic tissue lymphocyte subpopulations in liver disease. Scand J Gastroenteroll982: 17:855-60 36 Berman JS, Beer DJ, Theodore AC, Kornfeld H, Bernardo J, Center DM . Lymphocyte recruitment to the lung. Am Rev Respir Dis 1990; 142:238-57 37 Reynolds NY, Merrill WW Pulmonary immunology: humoral and cellular immune responsiveness of the respiratory tract. In : Simmons DH, ed . Current Pulmonary. New York: Wile y, 1981:318-422 38 Bjamason I, Ward K, Peters TJ. The leaky gut of alcoholism: possible route of entry for toxic compounds. Lancet 1984; 1:17982 39 Colombel JF, Vaerman JP, Mesnard B, Nemeth J, Dive C , Rambaud JC . Jejunal immunoglobulin secretion in alcoholic patients with and without cirrhosis. J Hepatoll991: 12:145-49

CHEST I 101 121 FEBRUARY, 1992

473