p53 immunostaining is correlated with reduced survival and is not correlated with gene mutations in resected pulmonary large cell carcinomas

Malignancy of pulmonary large cell carcinomas (LCC) increases from classic LCC through LCC with neuroendocrine morphology (LCCNM) to large cell neuroendocrine carcinomas (LCNEC). However, the histological classification has sometimes proved to be difficult. Because the malignancy of LCC is highly dependent on proteins with functions in the cell cycle, DNA repair, and apoptosis, p53 has been targeted as a potentially useful biological marker. p53 mutations in lung cancers have been shown to result in expression and protein expression also occurs in the absence of mutations. To validate the importance of both p53 protein expression (by immunostaining) and p53 gene mutations in lung LCC (by PCR-single strand conformational polymorphism analysis of exons 5, 6, 7, and 8) and to study their relationships with clinical factors and sub-classification we investigated the correlation of p53 abnormalities in 15 patients with LCC (5 classic LCC, 5 LCNEC, and 5 LCCNM) who had undergone resection with curative intent. Of these patients, 5/15 expressed p53 and none had mutant p53 sequences. There was a negative survival correlation with positive p53 immunostaining (P = 0.05). After adjustment for stage, age, gender, chemotherapy, radiotherapy, and histological subtypes by multivariate analysis, p53 expression had an independent impact on survival. The present study indicates that p53 assessment may provide an objective marker for the prognosis of LCC irrespective of morphological variants and suggests that p53 expression is important for outcome prediction in patients with the early stages of LCC. The results reported here should be considered to be initial results because tumors from only 15 patients were studied: 5 each from LCC, LCNEC and LCCNM. This was due to the rarity of these specific diseases.

Familial aggregation of cancer in patients with bronchogenic carcinoma.

History of cancer among first degree relatives was obtained in 124 patients with bronchogenic carcinoma [probands] and 248 controls, and differences in familial aggregation evaluated by calculation of odds ratio [OR] and their 95 percent confidence intervals [95% CI]. Probands were more likely than controls to have relatives with cancers [OR 1.27, 95% CI 0.49-3.10], both among smokers and nonsmokers [OR 1.31, 95% CI 0.27-6.79 and OR 1.21, 95% CI 0.12-6.16 respectively]. Sisters of probands were particularly at a higher risk [OR 8.72, 95% CI 0.85-430.08]. A genetic component, possibly independent of smoking habits, may be important in the causation of lung cancer.