[Cause of death, survival, and prognostic factors in a series of 98 patients with chronic myeloid leukemia]

Abstract

Purpose: To analyse the survival from diagnosis to blastic crisis, as well as the causes of death and the significance of different features in patients diagnosed of chronic myelogenous leukaemia (CML).

Material and methods: The series includes 98 patients who died of CML in Asturias Central Hospital and other nearby hospital between 1970 and 1990. The CML diagnosis had been established according to conventional criteria. The blastic crisis was diagnosed in accordance with morphologic and cytochemical studies, complemented in some cases with the assay of TdT or other markers. The Kaplan-Meier curves and long-rank test were applied to the statistical analysis of survival. The prognostic factors were evaluated by univariate correlation.

Results: The series' mean age was 48.57 years (range: 6-90) and the M/F ratio was 56/42. Cytogenetic study had been performed in 35 patients, of whom 33 had the Ph' chromosome. Death occurred in the blastic crisis in 67 cases (72%), in the accelerated phase in 14 cases (15%) and in the chronic phase in 12 cases (13%). The cause of death could be determined in 47 cases; of them, 24 (51%) died of infection, 11 (23.4%) of haemorrhage, and the remaining 12 (25.6%) of different complications. The median survival of the group as a whole was 32 months (range: 4-137). For the blastic crisis this figure was reduced to 2.5 months (range: 1-14), the lymphoid forms doing better, 5 months, than the non-lymphoid ones; 1.5 months (p < 0.03). Out of the data evaluated at diagnosis, only haemoglobin and the percentage of blast cells in peripheral blood showed any correlation with survival (r = +0.28, p < 0.05 for haemoglobin, and r = -0.30, p < 0.01 for blast cells).

Conclusions: (1) Although most patients die from infection or haemorrhage during the blast crisis, CML increases the risk of death in the accelerated and chronic stages as well. (2) Non-lymphoid blastic crisis has poorer prognosis than the lymphoid form. (3) The concentration of blast cells and haemoglobin in peripheral blood at diagnosis acquire prognostic significance in the present study.