Event models for tumor classification with SAGE gene expression data

Serial Analysis of Gene Expression (SAGE) is a relatively new method for monitoring gene expression levels and is expected to contribute significantly to the progress in cancer treatment by enabling a precise and early diagnosis. A promising application of SAGE gene expression data is classification of tumors. In this paper, we build three event models (the multivariate Bernoulli model, the multinomial model and the normalized multinomial model) for SAGE data classification. Both binary classification and multicategory classification are investigated. Experiments on two SAGE datasets show that the multivariate Bernoulli model performs well with small feature sizes, but the multinomial performs better at large feature sizes, while the normalized multinomial performs well with medium feature sizes. The multinomial achieves the highest overall accuracy.

Survival in 176 patients with inoperable non small cell lung cancer treated initially with chemotherapy in a district general hospital

In advanced non-small cell lung cancer (NSCLC) platinum based chemotherapy with second generation drugs improves median survival (MS) to 8 months and 29% and 10% at 1 and 2 years. Platinum with a third generation drug can improve survival further (BMJ 1995;311: 899) (Spiro et al. Thorax 2004;59:828 Big Lung Trial; N Engl J Med 2003;346:92 ECOG study). NICE now recommends chemotherapy with platinum and a third generation drug for inoperable NSCLC as the first treatment modality. Methods: We audited survival of 176/461 consecutive patients referred for at least 3 courses of platinum and either gemcitabine or vinorelbine from July 2001 to December 2005. Minimal follow up 17 months. Chemotherapy was given on site. Radical radiotherapy for stage IIIA, palliative radiotherapy and second line drugs were given as felt appropriate. Results: 64% were male. 30 (17%) were <55 years ; 66 (37.5%) age 55–65 years; 63 (35.8%) aged 66–75 and 16 (9.1%) >75 years. 5 (2.8%) were stage II; 46 (26%) stage IIIA; 68 (38%) stage IIIB and 55 (30.8%) stage IV. 68 (38%) had 0– 2 courses; 63 (36%) 3 courses and 44 (25%) had 4 or more.