Prognosis in acute myeloid leukemia and influence of monocytic markers : Epidemiological, clinical and experimental studies

Sammanfattning: The aim of the first part of this thesis was to describe an unselected series of AML patients in adults ( > 15 years) from the three counties Orebro, Värmland and Södermanland 1987-1992, including cases un-referred from district hospitals. In the study of incidence and registration, also ALL and unspecified AL were included due to delimitation difficulties. Four registries were used for case finding and diagnoses were verified from medical records. 260 cases of acute leukemias were found giving a mean annual incidence of 6.5 / 105 in adults. Under-notification, diversity and lack of specificity of codes, and selection of good prognosis patients were problems noted for the Cancer Registry. Supplementation with cases from the Cause of Death Registry substantially improved the coverage. The incidence of AML in adults was 5.4 / 105 / year and the median age 69.5 years. 56% had received standard induction treatment, 28% low-dose treatment and 16% no cytostatic treatment. Median survival for all patients was 5.8 months and the probability of survival at 5 years was 9.3%. In a multivariate Cox proportional hazard analysis, age, LDH and kidney function were independent prognostic variables for survival. Cytogenetics was missing in many patients but had a strong influence on outcome in univariate analysis. The aim of the second part was to investigate whether the "monocytic markers" ANAE, CD14 and lysozyme correlate to prognosis in AML. In a study of 65 AML patients, there was a marked difference in CD14 and ANAE positivity in bone marrow cells for some cases. Patients expressing 10% or more CD14 had a poorer prognosis, in the total material as well as when non-monocytic AML was analyzed separately. In a retrospective study of 232 AML patients treated in Orebro and Umeå, a linear relation was rejected and a bimodal relation was found between serum lysozyme and prognosis. Lysozyme values below 20 or above 80 mg/l were indicative of better outcome than values in the range 20-80 mg/l. The bimodal relation to prognosis was verified in multivariate logistic regression and Cox proportional hazard analyses with inclusion of the factors age, de novo / secondary AML, cytogenetic risk group, lysozyme, leukocyte count, lactic dehydrogenase (LDH), and kidney function. Lysozyme was a selected prognostic factor for CR frequency (p=0.0003), overall survival (p<0.0001) and CR duration (p=0.0005). The hazard ratios (HR) for lysozyme <20, 20-80 and >80 mg/l regarding overall survival were 1.0, 3.3 (95% Cl 2.1-5.1) and 0.7 (95% Cl 0.4-1.2). A bimodal relation to prognosis was found both in non-monocytic and monocytic AML. Our finding should be regarded as a new hypothesis and controlled in other studies. The aim of the third part was to investigate effects of lysozyme on the leukemia cell line K562. Hen egg white lysozyme (HEL) concentrations of 800-12800 mg/I gave increased proliferation with 1.4-2 fold higher cell counts and a tendency to higher viability compared to control incubations. HEL increased the proportion of cells in S-phase and decreased the proportion in G2/M-phase with an associated broadening of the peaks and higher fluorescence intensity signals in the DNA histograms. The threshold for effects of lysozyme was high, but corresponding levels of endogenous lysozyme are probably present in near cell environments of a not negligible proportion of patients with AML.