RESIDUAL DISEASE DETECTION IN MULTIPLE FOLLOW-UP SAMPLES IN CHILDREN WITH ACUTE LYMPHOBLASTIC-LEUKEMIA

The polymerase chain reaction (PCR) is more sensitive than other metho ds for detecting residual leukemic cells, and it can be used to analyz e the dynamics of the leukemic cell population during and after chemot herapy. Leukemia cells are identified among the normal cells through t he use of a leukemic cell-specific signature, such as the junction for med following rearrangement of the immunoglobulin (Ig) or T-cell recep tor (TCR) variable (V), diversity (D), and joining (J) regions. We clo ned the rearranged Ig or TCR genes from leukemic cells at the time of diagnosis by regular recombinant DNA techniques, or following PCR ampl ification, and determined the leukemic cell-specific V(D)J junctions. Using oligonucleotides prepared to the unique junctional sequences, we have analyzed Southern blots of PCR-amplified Ig or TCR genes from 11 patients at the time of diagnosis, and in from three to 14 follow-up samples for periods ranging from 17 to 97 months. The level of detecti on of residual blast cells in these patients was usually on the order of 1 leukemic cell in 10(4) normal ones. Conditions for the hybridizat ion and washing were individually adjusted so that the specificity of the probes was excellent. No residual leukemic cells were found in pat ients who remained in continuous complete remission at time points mor e than 6 months post-diagnosis. PCR detected impending relapse in four out of six cases, with PCR-detectable leukemic clones being present 2 , 4, 13 and 20 months prior to clinical signs of the disease. Relapse occurred in two patients who lacked PCR evidence of leukemic blasts in marrow samples collected 2 and 6 months before clinical recurrence, r espectively. The leukemic cells that persisted for 3 years after a rel apse in a patient who remained in clinical remission during this time period eventually became undetectable by PCR.