All-trans-retinoic acid (ATRA) has been incorporated in front-line therapy for newly diagnosed acute promyelocytic leukemia (APL). We conducted a multicenter study of differentiation therapy with ATRA alone or in combination with chemotherapy followed by intensive postremission chemotherapy in patients with APL (the JALSG APL92 study), and analyzed prognostic factors to increase the cure rate in our subsequent trial. From 1992 to 1997, adult patients with newly diagnosed APL received oral ATRA 45 mg/m2 daily alone until complete remission (CR) if initial leukocyte counts were < 3.0x10(9)/l, and ATRA daily plus daunorubicin (DNR) 40 mg/m2x3 days plus enocitabine (BHAC) 200 mg/m2x5 days if leukocyte counts were > or =3.0 x 10(9)/l. If peripheral blasts exceeded 1.0x10(9)/l during therapy, DNRx3 days plus BHACx5 days was added. After CR was achieved, three courses of consolidation and six courses of maintenance/intensification chemotherapy were administered. Of 376 patients enrolled, 369 were evaluable (median age 46 years, range 15-86 years; median leukocyte counts 2.0x10(9)/l), and 333 (90%) achieved CR (94% of patients treated with ATRA alone, 88% with ATRA plus later chemotherapy, 89% with ATRA plus initial chemotherapy, and 86% with ATRA plus initial and later chemotherapy). At a median follow-up of 45 months, the predicted 6-year overall and event-free survival (EFS) rates for all patients were 65% and 52%, respectively. Favorable prognostic factors for CR were younger age, no or mild purpura, high serum total protein level, low lactate dehydrogenase level, and no or mild disseminated intravascular coagulation (DIC). Favorable prognostic factors for EFS were leukocyte counts < 10.0x10(9)/l, mild DIC, and no sepsis during induction therapy. In the JALSG APL97 study, we intensified chemotherapy for patients with leukocyte counts > or =3.0x10(9)/l, and are randomly testing whether further chemotherapy is required for APL patients with negative PCR for PML/retinoic acid receptor alpha in the maintenance phase.