Reliability of bronchoalveolar lavage in the routine clinical assessment of patients with sarcoidosis. A retrospective analysis.

BACKGROUND

The aim of the study was to evaluate the reliability of bronchoalveolar lavage in the routine clinical assessment of patients with sarcoidosis.

STUDY DESIGN

retrospective analysis.

SETTING

2 Divisions of Respiratory Medicine, both in the context of a public general hospital (Ospedale Morelli, Sondalo, Italy and Ospedale Civile, Piacenza, Italy).

PATIENTS

89 patients with definitely diagnosed diffuse infiltrative lung disease, who underwent bronchoalveolar lavage as part of their initial diagnostic work-up in the period 1991-2001. Fifty-one of them had biopsy-proven sarcoidosis. Diagnostic procedures were: transbronchial lung biopsy (n=26), bronchial biopsy (n=11), skin biopsy (n=6), surgical lung biopsy (n=4), superficial lymph node biopsy (n=3), mediastinal lymph node biopsy (n=1). In 38 patients various interstitial lung diseases were diagnosed, according to established criteria.

MEASURES

we compared the percentage of BAL lymphocytes and CD4/CD8 ratio in the 2 groups. Sensitivity, specificity, positive and negative predictive value of BAL parameters were calculated with a cut-off value of 28% for BAL lymphocytes and 3.5 for CD4/CD8 ratio.

RESULTS

The percentage of lymphocytes was similar in the 2 groups (p=0.19); CD4/CD8 ratio was 5.7+/-4.9 in sarcoidosis and 2.0+/-3.3 in the other patients (p=0.0001). BAL CD4/CD8 ratio in sarcoidosis had sensitivity 58%, specificity 86%, positive predictive value 85%, negative predictive value 61%.

CONCLUSIONS

BAL has a low sensitivity in patients with sarcoidosis; CD4/CD8 ratio >or=3.5 strongly suggests sarcoidosis but is not specific enough if considered alone.