[In vitro drug resistance as independent prognostic factor in the study COALL-O5-92 Treatment of childhood acute lymphoblastic leukemia; two-tiered classification of treatments based on accepted risk criteria and drug sensitivity profiles in study COALL-06-97].

BACKGROUND Risk-adapted therapy in acute lymphoblastic leukemia (ALL) of childhood relies on traditional risk factors such as age, white blood count, immunological subtype, chromosomal aberrations and response to treatment. In spite of risk-adapted therapy, however, 20-30% of the patients suffer a relapse and may have profited from more intensive therapy. On the other hand a third of the patients is probably overtreated considering cure rates of 30% with less intensive therapies in the seventies. Additional prognostic criteria are therefore urgently needed. In a retrospective Dutch study in childhood ALL drug sensitivity testing with the MTT assay was identified as a new highly significant prognostic factor. PATIENTS AND METHODS In study COALL-92 in 140 patients in vitro drug sensitivity was performed on initial bone marrow or blood samples and as in the Dutch study a score was derived from the sensitivity of the 3 drugs prednisolone, vincristine and asparaginase (PVA score). For each drug a score of 1 (highest) to 3 (lowest sensitivity) is given. A score of 3 therefore indicates the best, a score of 9 the worst sensitivity. RESULTS Probability of event-free survival (pEFS) according to the PVA score was 0.94 for score 3 + 4, 0.80 for score 5-7, 0.35 for Score 8 + 9 (0.47 for score 7-9). For analysis within the low risk (LR) group (age 1-10 years, WBC < 25/nl, common or pre-B-ALL, remission day 28, no translocation 9;22 or 4;11) and high risk (HR) group 252 patients of the Dutch study and the COALL study were combined. In the LR group pEFS was 1.00 for score 3 + 4, 0.76 for score 5-7, 0.38 for score 8 + 9 (0.54 for score 7-9); in the HR group pEFS was 0.91 for score 3 + 4, 0.69 for score 5-7, 0.45 for score 8 + 9 (0.56 for score 7-9). Multivariate analysis identified the PVA score as independent prognostic factor. NEW STUDY COALL-97: In the new study COALL-97 patients are first stratified according to HR and LR criteria and then are stratified again according to the PVA score. LR and HR patients with a score of 3 + 4 receive less intensive treatment in reinduction therapy to reduce the cumulative dose of anthracyclines and to mitigate the high rate of infectious complications during this part of the protocol. LR patients with a score of 7-9 are treated according to the HR protocol; HR patients with a score of 8 + 9 receive BMT if an HLA identical family donor is available. In study COALL-97 the results of the minimal residual disease study and day 15 bone marrow will be compared with the PVA score. CONCLUSIONS In vitro drug sensitivity testing is an independent prognostic factor which allows adjustment of therapy to the individual risk of relapse in addition to the traditional risk factors. It can be assumed that in patients with a favourable resistance profile therapy can be reduced without loss of efficacy and that patients with an unfavourable resistance profile might profit from more intensive therapy.