An analysis of poor risk assignment in patients with germ cell tumours

Four methods for assigning patients with germ cell tumours to poor risk programmes were compared using 118 consecutive patients treated on 'good' and 'poor' risk studies. The median survival of patients deemed poor risk by Memorial Sloan-Kettering Cancer Center was 11 months, by Indiana University 15 months, by the National Cancer Institute 15 months, and by the European Organization for Research on the Treatment of Cancer 23.5 months. The major differences between these criteria were in the use of pretreatment tumour marker values, and the use of specific sites of metastases as independent prognostic variables. These data imply that a variable number of good prognosis patients enter some poor risk trials, the number being dependent upon the criteria used. Poor risk criteria should be as restrictive as possible in order to avoid over treating good risk patients, and to avoid artificially inflated response rates resulting from the successful treatment of such poor risk patients.