RESUMO
The real need for extensive staging at the time of diagnosis is discussed in regard to small cell lung carcinoma. We performed a decisional retrospective analysis on a series of 182 patients, based on three staging steps: the first step included physical examination and routine biologic tests. The second step consisted of liver ultrasonography and needle aspiration of any clinically detectable tumor mass, and the third step included bone marrow examination, radionuclide bone scan, thoracic, abdominal, and brain CT scan. A stepwise multivariate logistic regression performed on 11 variables considered in the first step shows that a four-parameter model can predict the spread of the disease (limited or extensive): weight loss, performance status, and elevated LDH or alkaline phosphatase levels. Limited disease can be predicted in two ways: (1) elevated LDH with normal alkaline phosphatases, no weight loss, and good performance status, or (2) normal LDH and alkaline phosphatases. In this series, 28 percent of patients can be predicted as having extensive disease and can be treated with chemotherapy alone without chest irradiation. After the second step, the probability of disease being extensive is only 25 percent, and only 84 (46.15 percent) patients would need to undergo the third step of staging procedures (brain CT scan, bone marrow aspiration and biopsy, radionuclide bone scan) with this method. We conclude that a multistep approach represents a simple staging method and offers the advantage of harmlessness and lower costs for patients not to be evaluated in prospective clinical trials.