Classification and prognostic variables in myelomatosis

Abstract

It is difficult to compare the results of treatment obtained in different trials in myelomatosis because different sets of diagnostic criteria are used, and because the criteria by which patients are deemed eligible for entry vary. Thus the composition of different series of patients varies considerably. Furthermore, the outcome of treatment is recorded in different ways. Uniformity in the diagnostic categories entered would reduce the variance in survival between different trials: for example, trials in myelomatosis should exclude patients with monoclonal gammopathy of uncertain significance, non-progressive or indolent myeloma, extramedullary plasmacytoma, and plasma-cell leukaemia. The subdivision into simple prognostic groupings such as those proposed by the Medical Research Council is helpful in interpreting the survival patterns in different trials in which the proportions of patients in different prognostic groups are likely to vary. These groupings and other staging systems do not correlate with responsiveness to treatment. Rapid responders fare worse than slow responders, and this might provide a basis for a second randomisation to test whether a change in treatment could benefit the rapid responders.