Methodological aspects of a clinical trial in chronic lymphocytic leukemia: necessity to develop new biostatistical...

The CLL protocol allowed a better understanding of the natural history of Chronic Lymphocytic Leukemia and an assessment of several treatment strategies. From the point of view of biostatistical research, it raised several important problems. In this paper, we show the interest of applying sequential methods, particularly the triangular test. Other biostatistical problems raised by the CLL 80 protocol are still not, or only partially, resolved. They include taking into account prior belief about treatment benefit in sample size calculations, using surrogate endpoints, developing methods to take into account the causes of death, to analyse repeated measurements, and to assess and compare staging systems.