LIKELIHOOD RATIO FOR PULMONARY SARCOIDOSIS AS A FUNCTION OF CD4/CD8 RATIO IN BRONCHOALVEOLAR LAVAGE FLUID

Dieter De Smet, L. Verfaillie, Wim Renmans, O. Heylen, G. Martens, Marc De Waele

Research output: Contribution to journalConference paper

Abstract

BACKGROUND
Sarcoidosis is a chronic inflammatory disorder of unkown origin with most frequently affects the lungs. Noncaseating granulomas are the key pathologic finding which makes a bronchial biopsy obligatory to confirm the diagnosis of sarcoidosis. Since and elevated CD4/CD8 ratio of BALF lymphocytes is generally considered to suport the diagnosis of pulmonary sarcoidosis (PS), flowcytometric immunophenotyping of bronchoalveolar lavage fluid (BALF) lymphocytes has become and important diagnostic tool. However, an elevated BALF CD4/CD8 ratio by itself lacks specificity and sensitivity; and moreover, studies that addressed the clinical usefulness of this ratio used varying cut-of values (2.5 to 3.5)which were chosen along different strategies.
AIM
In this study we examined whether calculating likehood ratios (LRs) for PS different BALF CD4/CD8 ratio intervals can contribute to the clinical interpretation of these results.
MATERIALS AND METHODS
We examined BALFs of 374 patients with pulmonary disease; these included 60 patients with biopsy-confirmed PS and 314 patients with other lung disorders such as asthma, COPD, pneumonia, neoplasms, tuberculosis and other interstitial lung disorders. The CD4/CD8 was calculated after flowcytometric enumeration of lymphocyte subsets according to our laboratory's standard operating procedures.
RESULTS
A significantly higher CD4/CD8 ratio in BALF was found for PS patients, as compared to non-PS patients (p7.5. For patients with PS, the likelihoods were respectively 0%, 18.3%,41.7%, 18.3% and 21.7%. For the non-PS patients, the values we were respectively 14.3%, 45.2%, 36.9%, 3.5% and 0%. Next, we calculated the LR for each test result interval by dividing the corresponding likelihoods. The LRs (95% Cl) were respectively 0 (0.000 to 0.931), 0.41 (0.234 to 0.701), 1.13 (0.809 to 1.573), 5.23 (2.378 to 11.517) and approaching infinity (8.183 to infinity).
DISCUSSION
Our data show that a BALF CD4/CD8 ratio between 15 and 5.0 has almost no clinical usefulness for excluding or supporting diagnosis of PS. This is in contrast to the frequently described cut-off values of 2.5 or 3.5. On the other hand, our results show that a CD4/CD8 ratio of less than 0.5 or more than 7.5 respectively excludes or confirms the diagnosis. The other result intervals (0.5-1.5 and 5-7.5) carry valuable but more limited diagnostic information.
CONCLUSION
We have illustrated how the use of LRs for different test result intervals can improve clincal usefulness of BALF CD4/CD8 ratio assessment for PS. Clinical laboratories might consider providing such LRs assist clinical interpretation.
Original languageEnglish
Pages (from-to)218-218
Number of pages1
JournalActa Clinica Belgica
Volume65
Issue number2010
Publication statusPublished - 1 May 2010

Keywords

  • PULMONARY SARCOIDOSIS
  • CD4/CD8 RATIO
  • BAL fluid

Fingerprint

Dive into the research topics of 'LIKELIHOOD RATIO FOR PULMONARY SARCOIDOSIS AS A FUNCTION OF CD4/CD8 RATIO IN BRONCHOALVEOLAR LAVAGE FLUID'. Together they form a unique fingerprint.

Cite this