Purpose: Our aim in this study was to compare prognostic models based on laboratory tests with a model including imaging information in small-cell lung cancer.
Patients and methods: A retrospective analysis was performed on 156 consecutive patients. Three existing models based on laboratory tests and performance status (PS) and a model based on disease stage assessed by imaging techniques and PS were tested with Cox regression analysis.
Results: The 3 laboratory-based models and the imaging-based model were significant in predicting prognosis in our patient group, with hazard ratios of 1.6-3 for medium prognosis groups and 2.6-6.1 for poor prognosis groups compared with good prognosis groups. Models based on laboratory tests appear to predict survival probabilities at least as well as a model with information from imaging techniques.
Conclusion: Prognostic models using PS and laboratory tests provide a similar estimation of survival of patients with small-cell lung cancer as the combination of PS and disease stage assessed by imaging tests.