BACKGROUND AND OBJECTIVE: In a prior study, primary resistant acute lymphoblastic leukemia (RES-ALL) was observed in 11 of 176 (6%) adult patients treated with a four drug regimen (IVAP), its incidence being higher in T-cell or Philadelphia (Ph) chromosome/BCR-ABL rearrangement positive ALL cases with a blast cell count >25x10(9)/L (RES-ALL rate 19%, p=0.04). Aiming to minimize this percentage of resistant disease, fractionated cyclophosphamide (f-CY) was then added to the IVAP regimen. DESIGN AND METHODS: Study 08-96 was a prospective, collaborative phase II trial carried out at eight general hospital centers specialized in the care of hematologic malignancies. Historical IVAP-treated patients served as a retrospective control group. All consecutive, untreated patients (>15 years) with a diagnosis of ALL or advanced-stage lymphoblastic lymphoma (LBL) were eligible. RES-ALL was defined as the persistence of >5% ALL cells in the bone marrow 28-40 days after the start of the IVAP regimen (idarubicin 10 mg/m(2)/d on days 1 and 2; vincristine 2 mg on days 1, 8 and 15; L-asparaginase 6,000 U/m(2) on alternate days 3 6 from day 8; prednisone 60 mg/m(2)/d on days 1-21). In the new study, two f-CY schedules were sequentially adopted: CY 150 or 75 mg/m(2)/bd, given for 4 consecutive days before IVAP (f-CY 1200 or 600, expressing total CY dose in mg/m(2)). RESULTS: Eighty-eight patients were evaluable (age range 15-74 years, blast count 0-240x10(9)/L, 14 T-lineage, 74 B-lineage, 13 Ph/BCR-ABL+). The first 39 patients received the f-CY 1200 schedule, 22 patients received f-CY 600, and the last 27 patients were not given any f-CY. These changes were dictated by the results of interim analyses of the f-CY groups (RES-ALL rate not reduced, myelotoxicity increased). Altogether, compared with the historical IVAP and no f-CY groups, the incidence of RES-ALL was not decreased by the addition of f-CY 1200/600 in B-lineage ALL, regardless of Ph/BCR-ABL expression and blast count. However, none of 14 T-ALL cases in the new study had RES-ALL (8 in f-CY groups, 5 of whom with >25x10(9)/L blast cells), compared to 5/39 (13%, overall) or 4/21 (19%, with >25x10(9)/L blast cells) among the control cases. Owing to small sample size, this difference was not statistically significant. INTERPRETATION AND CONCLUSIONS: This preliminary experience suggests that T-ALL may be more sensitive than B-lineage ALL to an early therapy including f-CY. The hypothesis could be tested in a larger clinical trial.