Background: Minimal residual disease (MRD) detected during remission might predict outcome in children with acute lymphoblastic leukemia. No population‐based studies have been carried out. We studied all children with ALL presenting over 5 years within a defined population to determine its clinical importance.
Procedure: Patients were investigated for the presence of unique clonal rearrangements of IgH and T‐cell receptor genes. Unique patient specific probes were used to detect, by polymerase chain reaction, the presence of clonal markers indicating MRD in mononuclear cells obtained from marrow samples at 1, 3, 5, and 24 months. The effect of MRD on event‐free survival was determined.
Results: Seventy‐seven of 120 children with ALL had informative markers and samples of remission marrow suitable for testing. Presence or absence of MRD did not significantly affect outcome. Gender (P < 0.04) and white cell count (P < .04) were independent prognostic factors. Analysis of only those cases with detectable MRD showed that cases with one blast per 100 mononuclear cells, or more, 28 days after starting treatment did worse than those with lower levels (hazard ratio 7.77, P < 0.02).
Conclusions: Mere presence or absence of MRD is probably too crude a measure to be useful and worse than other standard prognostic indicators. A threshold of 10−2 blasts at 28 days might be discriminatory, but should not be over‐interpreted. The number of patients available for this analysis (31) was small, the threshold and sampling points were arbitrary and any effects could be treatment regimen‐specific. Large prospective studies are needed.