The management of chronic lymphocytic leukemia at a single centre over a 24-year period: prognostic factors for survival.

Over a 24-year period, 137 patients were referred for management of newly diagnosed chronic lymphocytic leukemia. One hundred and nineteen patients have been reviewed in terms of response to therapy and prognostic factors for survival; 18 patients were excluded either because lymph node biopsy was not compatible with the diagnosis of CLL (11 patients), or because the lymphocyte count at presentation was < 5 x 10(9)/l (seven patients). Patients were staged retrospectively according to both the Rai and Binet Classifications. Forty-eight per cent (57/119) were deemed not to be in need of any treatment at presentation, 36 per cent (43/119) have never received any specific therapy. The majority of patients received chlorambucil alone, at a dose of 10 mg daily given for 6 weeks, followed by a 2-week interval, followed by three, 2-week cycles. The overall response rate (complete+partial remission) was 38 per cent. In terms of survival, there was a trend in favour of patients who responded to treatment in comparison with those who did not but this did not reach statistical significance (P = 0.07). Correlations with stage were highly significant, the median survivals for patients with stage A, B and C disease (Binet) were 12.5, 8 and 3.5 years respectively. On univariate analysis, the absolute lymphocyte count at presentation was the most significant prognostic factor for survival, patients presenting with an absolute lymphocyte count above 50 x 10(9)/l having a less favourable prognosis (P = 0.002). However, on multivariate analysis, older age, a low hemoglobin, low platelet count, and the presence of lymphadenopathy and fever at presentation correlated adversely with survival. Overall, 40 patients died as a consequence of CLL or from disease-related causes, 34/40 dying of infection. Twenty-one patients developed second cancers. With a median follow-up of 13 years, these results confirm that the two staging systems can separate patients into prognostic groups, however in practice, there is heterogeneity of outcome within stage. New approaches are urgently needed.