We entered 26 patients with newly diagnosed acute promyelocytic leukemia (APL) in a pilot study of all-transretinoic acid (ATRA) followed by intensive chemotherapy. Median age was 46 (range 25 to 63). No patient presented with leukocytes > 10 x 10(9)/L or had the microgranular APL variant. Cytogenetic analysis (25 patients) found a t(15;17) in 24 cases. Patients were scheduled to receive ATRA (45 mg/m2/d) until complete remission, followed by an intensive daunorubicin (DNR) + Ara C course ("4 + 7" course), then three "2 + 5" DNR + Ara C courses and maintenance chemotheapy. However, the "4 + 7" course was administered in emergency if hyperleukocytosis rapidly developed to prevent leukostasis. Twenty-five patients (96%) achieved CR, 14 with ATRA alone and 11 after the addition of the "4 + 7" course on day 2 to 30 of treatment, because leukocytes rapidly increased (9 cases), because of resistance to ATRA (1 case), and development of organomegaly (1 case). The remaining patient died on day 6, from CNS bleeding. Apart from hyperleukocytosis, side effects were usually moderate. In the 11 patients who could be studied in vitro, a very good correlation was found between in vivo and vitro differentiation and proliferation of APL blasts with ATRA. Three patients were allografted after the "4 + 7" course. Four patients did not receive this course but received the subsequent "2 + 5" courses and maintenance. The remaining patients followed the scheduled protocol. Three patients relapsed after 8, 11, and 15 months (including one allografted patient). Two patients died in CR, after 6 and 17 months. The other 20 patients remained in CR after 18+ to 34+ months (median 21). Actuarial disease free interval (DFI) and event free survival (EFS) were 87% and 77%, respectively, after 18 months. These results were compared to those obtained in our previous APL 84 trial with chemotherapy alone in newly diagnosed APL (after excluding patients included in this trial who presented with hyperleukocytosis). In APL 84 trial, the CR rate was 76%, the actuarial DFI and EFS were 59% and 48% after 18 months, respectively. Differences with the pilot study of ATRA followed by chemotherapy were significant for DFI (P = .02), EFS (P = .006), but not for CR rate (P = .08). Although this is a historical comparison, these results suggest that ATRA followed by chemotherapy may prove superior to chemotherapy alone in newly diagnosed APL, by slightly increasing the CR rate, but perhaps more importantly by reducing the relapse rate.(ABSTRACT TRUNCATED AT 400 WORDS)