SUBJECTS

Seventeen consecutive patients with newly diagnosed biopsy proven pulmonary sarcoidosis were recruited from the Respiratory Unit of King Edward VIII Hospital, a major teaching hospital in Durban, South Africa. Of these, 14 patients (seven men) of mean (SD) age 40.6 (1.8) years were able to produce an adequate sputum sample at baseline and following six months of treatment and comprised the study cohort. Patients were treated with an initial dose of prednisone of 40 mg/day orally for six weeks, tapering to 10 mg/day within six months.

Healthy controls comprised 12 non-smoking, non-atopic volunteers of mean (SD) age 41 (2.8) years, matched for age, sex, and race who were able to produce adequate sputum samples following sputum induction. They had had no respiratory tract illnesses in the six weeks preceding the study. Their forced expiratory volume in one second (FEV₁) was 98 (8)% predicted.

STUDY PROTOCOL

The ethics committee of the University of Natal approved the study and signed informed consent was obtained from all patients before entry into the study. Spirometric tests, serum angiotensin converting enzyme levels, liver function, urea and electrolytes, full blood count, serum calcium levels, 24 hour urinary calcium excretion, chest radiography, and high resolution computed tomographic (HRCT) scan of the chest was performed on all subjects. The HRCT scan was done as part of a larger study correlating lung function and radiological features in sarcoidosis with exhaled nitric oxide levels in these subjects. Induced sputum was obtained at baseline and four days later to test for reproducibility of the differential cell counts. BAL was performed four days later. Induced sputum and BAL were repeated six months after the baseline study.

SPUTUM INDUCTION

Sputum induction was performed according to the standard procedure described elsewhere.4 The subjects were instructed to rinse their mouth thoroughly with water after inhalation of 200 μg aerosolised salbutamol. They then inhaled 3% saline at room temperature, nebulised for 20 minutes via an ultrasonic nebuliser (DeVilbiss 99; Heston, UK) set at maximum output (4.5 ml/min).

Subjects were encouraged to cough at five minutes and then at three minute intervals into polypropylene pots. Saliva was initially expectorated into a separate container. Sputum plugs up to an approximate volume of 3 ml were selected.

PROCESSING OF SPUTUM SAMPLES

A 1 ml sample of sputum was spun at 60 000 rpm for 10 minutes and the supernatant was stored at –70°C for cytokine analysis. The rest of the sputum sample was diluted with 2 ml Hank's balanced salt solution (HBSS) containing 1% dithioreitol (DTT) (Sigma Chemicals, Poole, UK) and gently vortexed until homogenous. The cell pellet was resuspended in 2 ml HBSS. Total cell counts were determined by a haemocytometer using a Kimura stain. Slides were prepared using a cytospin (Shandon, Runcorn, UK) and stained with May-Grunwald-Giemsa for differential cell counts. Two observers blinded to the patients' characteristics counted 400 cells. Differential cell counts were expressed as a percentage of lower airway cells.

One ml of the suspension was processed for flow cytometry. The suspension of sputum cells in DTT was layered across a density gradient medium (Histopaque 1077, Sigma) and cytocentrifuged at 1000 rpm for 10 minutes. The cell pellet obtained was resuspended in 2 ml HBSS and incubated with 2 ml 10% bovine serum albumin at room temperature for 20 minutes to reduce non-specific binding. This suspension was then centrifuged at 3000 rpm for two minutes and the cell pellet was resuspended in HBSS. Pairs of monoclonal antibodies CD3/CD19 (T and B cells), CD3/CD4 (T helper cells), CD4/CD25 (activated helper T cells), CD8/CD25 (activated suppressor T cells), and CD3/CD8 (suppressor T cells) (Sigma Diagnostics, Poole, UK) were then added to the suspension and incubated for 30 minutes. The suspension was then subjected to flow cytometry to determine the CD4 and CD8 lymphocyte subsets.7

BRONCHOSCOPY AND BAL

An Olympus bronchoscope FIT20D (Tokyo, Japan) was inserted and wedged in the right middle lobe under adequate local anaesthesia and sedation with midazolam. A segment was lavaged using three 60 ml aliquots of saline. The BAL fluid samples were analysed for total and differential cell counts, TNFα levels, and flow cytometry to measure CD4:CD8 lymphocyte counts.

PROCESSING OF BAL FLUID

The pooled BAL fluid specimen was passed through sterile nylon gauze. Aliquots of 2 ml lavage fluid were cytocentrifuged at 1000 rpm for 10 minutes. The supernatant was stored for the measurement of TNFα levels and the cell pellet was resuspended in HBSS. A concentration of 1 × 10⁵ cells/ml was prepared and a cytospin slide preparation made (Shandon) and stained with May-Grunwald-Giemsa. Two observers blinded to the patients' characteristics counted 400 cells to determine the differential cell count.

A further 2 ml of lavage fluid was resuspended in HBSS and layered onto a density gradient medium (Histopague-1077, Sigma Diagnostics, UK) and cytocentrifuged at 600gfor 10 minutes. The cell pellet was then resuspended in HBSS and incubated with 10% bovine serum albumin (Sigma) for 20 minutes. As for sputum, pairs of monoclonal antibodies to CD4 and CD8 lymphocytes were then added to the suspension and incubated for 30 minutes after which flow cytometry was performed.

FLOW CYTOMETRY

Flow cytometry was performed using an Epic-Profile-II flow cytometer (Hialeah, Florida, USA) that detects lymphocytes by fluorescence. The fluorescence was backgated to show that lymphocytes in the sputum had similar physical properties (forward and side scatter, size and granularity) to lymphocytes in BAL fluid. Lymphocytes lie in a cigar-shaped cluster with a low wide scatter and more elongated forward scatter than lymphocytes in the blood. To enhance the number of lymphocytes for analysis, an acquisition gate was set using the lower third of the side scatter field. The analysis gate was broad in order to exclude non-lymphocytic cells. Results were expressed as percentage of the cells detected by fluorescence and not as a percentage of the gate because of contaminating debris in the sputum. A minimum of 400 lymphocytes detected on the CD3/CD19 panel was regarded as acceptable because it is the amount used in immunohistochemistry.7

TNFα ASSAY

The TNFα concentrations were measured using an enzyme linked immunosorbent assay (ELISA) kit obtained from Genzyme (Cambridge, Massachusetts, USA).4 Briefly, the kit contained 96-well microtitre plates precoated with bovine serum albumin. Samples were added to the wells and incubated with monclonal anti-TNFα antibodies for four hours. The optical density of the plate was then read using a plate photometer. The detection limit of the assay was 0.5 pg/ml.

STATISTICAL ANALYSIS

Cell counts were expressed ×10⁶/ml and differential cell counts were expressed as a percentage of the mean (SE). The intraclass correlation coefficient was used to evaluate the reproducibility of differential cell counts. Pearson's correlation coefficients were used to establish correlations between the individual cellularities, CD4:CD8 ratio, and TNFα levels in sputum and BAL fluid. An r value of >0.7 was regarded as acceptable and a p value of <0.05 was regarded as significant. The 95% confidence intervals (CI) for the CD4:CD8 ratios and TNFα levels are also presented.