Epidermal growth factor receptor, cyclooxygenase-2, and BAX expression in the primary non-small cell lung cancer and brain metastases.

PURPOSE: The purpose is to identify biological markers that predict brain metastasis and treatment outcome in non-small cell lung cancer (NSCLC). EXPERIMENTAL DESIGN: Samples were obtained from the primary tumors, lymph nodes, and brain metastases of 29 patients with NSCLC who had undergone resection of both the pulmonary tumors and the brain lesions. Samples from 29 patients matched for age, sex, and histology whose pulmonary tumors were resected served as controls. Samples were stained with H&E as well as immunohistochemical stains for epidermal growth factor receptor (EGFR), cyclooxygenase 2 (COX-2), and BAX. Comparisons were made between patients with and without brain metastasis. Independent investigators determined the percentage of positive cells. RESULTS: There was positive correlation in expression of all three biomarkers between primary lung tumors and lymph node metastases. Significantly higher levels of EGFR were found in lymph node metastases in the control group (P = 0.0147). COX-2 expression in brain lesions correlated with expression in primary tumors (P = 0.023). BAX levels were lower in poorly differentiated tumors in lymph node metastases in the control group (P = 0.01) and in brain metastases (P = 0.045). Low EGFR expression and high COX-2 expression in lymph node metastasis were associated with poorer treatment outcome. CONCLUSIONS: Expression of EGFR, COX-2, and BAX in primary lung tumors did not differ between patients with brain metastases from NSCLC and those without brain metastases. These three biomarkers cannot be used to predict brain metastasis. Studies of other biomarkers are under way in an effort to predict brain metastasis among patients with NSCLC.

Association of Ki-67, p53, and bcl-2 expression of the primary non-small-cell lung cancer lesion with brain metastatic lesion.

PURPOSE: The study was conducted to determine whether immunohistochemical analysis of Ki-67, p53, and bcl-2 in patients with non-small-cell lung cancer is associated with a higher rate of brain metastases and whether the intrapatient expression of these biomarkers (in the primary tumors vs. brain lesions) is similar. METHODS AND MATERIALS: At the M. D. Anderson Cancer Center, tumors from 29 case patients with primary lung tumor and brain metastasis and 29 control patients with primary lung tumor but no brain metastasis were resected and examined for immunohistochemical expression. Ki-67, p53, and bcl-2 were analyzed in resected primary lung, lymph node, and metastatic brain tumors. Each control patient was matched by age, gender, and histology to a patient with brain metastasis. RESULTS: No significant differences in patient survival characteristics were detected between the case group and control group. Also, difference in patient outcome between the two groups was not generally predicted by biomarker analysis. However, when the groups were combined, the biomarker analysis was predictive for certain patient outcome end points. Using median values as cutoff points between low and high expression of biomarkers, it was observed that high expression of Ki-67 (>40%) in lung primaries was associated with poorer disease-free survival (p = 0.04), whereas low expression of p53 in lung primaries was associated with poorer overall survival (p = 0.04), and these patients had a higher rate of nonbrain distant metastases (p = 0.02). The patients with brain metastases were particularly prone to developing nonbrain distant metastases if the percentage of p53-positive cells in brain metastases was low (p = 0.01). There was a positive correlation in the expression of Ki-67 (p = 0.02)(r(2) = 0.1608), as well as p53 (p < 0.001) (r(2) = 0.7380), between lung primaries and brain metastases. Compared to Ki-67 and p53, bcl-2 was the least predictive. CONCLUSION: Differences in biomarker expression between the case and control groups did not serve as significant predictors of brain metastasis or patient survival. There was a strong correlation between lung primary biomarker expression and brain metastasis expression for Ki-67 and p53. Univariate analysis showed that low p53 and high Ki-67 expression predicted poor prognosis. This study shows that there may be a strong correlation between biomarker expression in non-small-cell lung cancer primary tumors and their brain metastases.