PROGNOSIS OF PATIENTS WITH ADVANCED HODGKINS-DISEASE - EVALUATION OF 4 PROGNOSTIC MODELS USING 344 PATIENTS INCLUDED IN THE GROUP-DETUDES-DES-LYMPHOMES-DE-LADULTE STUDY

BACKGROUND, To determine whether a high risk group could be identified within a group of patients with advanced stage Hodgkin's disease (HD) , the authors applied several prognostic models to patients treated ac cording to the H89 protocol. METHODS. This study included 344 patients with Stage IIIB-IV HD who were treated with chemotherapy alone (8 cyc les) or chemotherapy (6 cycles) plus radiation therapy. Four prognosti c models were selected for this study: the numeric prognostic index of the Scotland and Newcastle Lymphoma Group, the Christie Hospital (Man chester)-St. Bartholomew's Hospital (London) model, the Memorial Sloan -Kettering Cancer Center (MSKCC) model, and the criteria used in the E uropean Bone Marrow Transplant (EBMT)/Intergroup Trial. RESULTS. Univa riate analysis of H89 protocol patients showed that 5 variables includ ed in the models had prognostic significance: age > 45 years (P = 0.00 01), anemia (hemoglobin < 12 g/dL for males and < 10 g/dL for females) (P = 0.0001), number of extranodal sites greater than or equal to 2 ( P = 0.0013), serum lactic acid dehydrogenase greater than the normal v alue (P = 0.0018), and lymphocyte count < 0.75 x 10(9)L(-1) (P = 0.006 3). All four models divided patients into prognostic subgroups. Signif icant differences among the subgroups were found by log rank analysis (chi-square test = 11-48; P = 0.01-0.0001). The worst prognostic group defined by the MSKCC model (greater than or equal to 3 adverse factor s) had an overall survival rate of 59% at 3 years and an estimated 3-y ear event free survival rate of 43%. CONCLUSIONS. Patients with at lea st three adverse factors according to the MSKCC model or the EBMT crit eria had a higher risk of failure with conventional treatment; however , based on survival rate, no very high risk group could be identified. Nonetheless, these prognostic models may be useful to recognize patie nts with good prognosis who can be cured with conventional therapy and for whom treatment morbidity and mortality can be minimized. (C) 1997 American Cancer Society.