[Prognostic factors and survival in chronic myeloproliferative disorders]

The broad spectrum of clinical and hematological as well as histomorphological findings at diagnosis of chronic myeloproliferative disorders (CMPDs) is significantly associated with prognosis. However, in this context the impact of bone marrow histology is still being controversially discussed. This feature applies to CML in particular. Therefore, risk classification is mainly based on clinical data. In order to evaluate the predictive value of bone marrow morphology we performed a retrospective study on a total of 1023 patients with CMPDs. Relative survival rates and a disease-specific loss of life expectancy were calculated to adjust the age- and gender-specific mortality in older patients. Patients with chronic myeloid leukemia (CML) showed an average life expectancy of 5 years, with significantly longer survival times under interferon treatment. In contrast, the ET group did not disclose any relevant reduction in life expectancy. Initial bone marrow fibrosis and a reduction in erythropoiesis were the most important prognostic features in CML. Furthermore, Pseudo-Gaucher cells indicated a favorable outcome in the cohort of patients receiving chemotherapy. On the other hand, peripheral myelo- and erythroblasts were correlated with a worsening in survival. Regarding IMF, a simplified multivariate risk score was constructed, including age, hemoglobin level, platelet and leukocyte counts, and a leuko-erythroblastic blood picture as most important variables. The three risk groups derived showed significantly different survival patterns, but in this calculation bone marrow histology exerted no major influence on survival. On the other hand, initial (prefibrotic) stages of IMF revealed a better prognosis. In conclusion, our results underline the importance of bone marrow morphology in CMPDs, since significant correlations with patients' outcome were calculated. Particularly in CML, myelofibrosis and reduction of erythropoiesis were associated with survival and, thus, must be regarded as important predictive and independent parameters.