A translational study to investigate the association between smoking-induced lung inflammation and lung metastases (LM) from breast cancer (BC)

10514 Background: Retrospective and preclinical studies suggest that smokers are at increased risk for LM from BC, possibly mediated by lung inflammation and the bioactive lipid, PGE2. To investigate this link in pts, we examined urinary PGE-M, a stable end metabolite of PGE2. Methods: A translational study was conducted, consisting of pts with LM (n=100), pts with mets but no known lung mets (NKL, n=100) and controls (CTRLS) with a history of BC (n=200). Pts receiving steroids and radiotherapy were excluded. Pts gave urine (PGE-M), blood (biomarkers), had BMI measured, and completed a validated questionnaire (smoking, NSAIDs, confounders). PGE-M was measured by mass spectrometry, normalized to urinary creatinine and compared between grps in both univariate and multivariate models, correcting for covariates. Results: From 09/2010 to 06/2011, 400 pts, med age 58 yrs (range 24-88) enrolled. There were no significant differences between grps in smoking and NSAID exposure (table). PGE-M (med; range) was highest among pts with LM (6.7; 0.7 - 43.4) compared to pts with NKL (4.6; 0.7 - 26.8) and CTRLS (4.2; 0.9 – 62.6, P24 pack year smoking history and LM had the highest co-variate-adjusted log PGE-M (P=0.03). In this model, age (P