LONG-TERM SURVIVAL IN SMALL-CELL LUNG-CANCER - THE CASE FOR A STANDARD DEFINITION

Objectives. This paper investigates whether a particular time point ca n be recommended as the standard definition for long-term survival (LT S) in small cell lung cancer (SCLC). Design. A review of the literatur e specifically referring to long-term survival in SCLC in the title. I ndividual and updated survival data from 2196 patients entered into si x Medical Research Council (MRC) SCLC randomised trials to investigate changes in the hazard, or risk of death, over time. Examination of su bgroups to identify different hazard patterns. Setting. World-wide lit erature and patients recruited to SCLC trials conducted by the MRC Lun g Cancer Working Party (LCWP). Subjects. Papers accessed through compu terised literature and hand searches and 2196 patients from six random ised clinical trials in SCLC conducted by the MRC LCWP. Results. In al l, 111 publications were identified by the literature searches. Althou gh the majority defined LTS as 2 years, the definitions ranged from 18 months to greater than 5 years. There thus appears to be no agreed st andard definition. The daily hazard was plotted for the large series o f patients entered into the MRC LCWP SCLC trials to observe any patter ns of change. There was an approximately constant daily hazard of 0.00 35 during the first 2 years from randomisation, and following a transi tional period, there was a further approximately constant, but smaller , daily hazard of 0.00035 from 3 years onwards. When subgroups of pati ents were examined this transition to a lower risk was observed in pat ients with limited disease, but patients with extensive disease remain ed at a high constant hazard throughout. It is suggested that 3 years should be adopted as the standard definition of LTS in SCLC, and that studies should always report limited and extensive stage patients sepa rately.