Abstract
Retinoic acid syndrome (RAS) can be a life-threatening complication in patients with acute promyelocytic leukemia (APL) undergoing induction therapy with all-trans retinoic acid (ATRA). Incidence of RAS has been reported ranging from 2% to 30%. It has been suggested that patients with leukocytes >5 x109/L at presentation are at high risk for the development of RAS. The impact of RAS on long term outcome is still a matter of controversy. We analyze the incidence, prognostic factors and outcome of RAS in 733 patients with newly diagnosed APL enrolled in the PETHEMA LPA96 and LPA99 trials (175 and 558 patients, respectively). Induction therapy consisted of ATRA and idarubicin, followed by three consolidation courses of anthracycline monochemotherapy. In the LPA99 trial, ATRA was added in each cycle of consolidation, except for low-risk patients. In the LPA99 trial, all patients received RAS prophylaxis with oral prednisone (0.5 mg/kg). Temporary discontinuation of ATRA and treatment with intravenous dexamethasone were recommended at the first signs of suspected RAS, in both trials. Definite RAS was defined as the presence of at least four of the following criteria: unexplained fever, respiratory distress, radiological pulmonary infiltrates, pericardial/pleural effusion, hypotension, renal failure, and weight gain over 5 kg. Overall, 87 patients (12%) experienced RAS, after a median of 6 days of ATRA (range, 0 to 46). Forty-seven cases (54%) occurred from days 0 to 7, 4 (5%) from days 8 to 14, 32 (36%) from days 15 to 30, and 4 (5%) from days 31 to 46. The main clinical signs were pulmonary infiltrates (83%), fever (80%), weight gain (74%), pleural effusion (63%) and renal failure (49%). ATRA was discontinued in 63% of patients. RAS was associated with age >50 years (41% vs 29%, p=0.02), serum level of creatinine >1.4 mg/dl (9% vs 3%, p<0.01) and leukocytes at presentation >5x109/L (46% vs 32%, p=0.01). Leukocytes >5x109/L and creatinine >1.4 mg/dl remained as independent prognostic factors in multivariate analysis. The incidence of RAS was not statistically different between the LPA96 (without prednisone prophylaxis) and LPA99 trials (15% vs 11%, p=0.16). RAS was associated with induction death (26% vs 7%, p<0.01) and was the main cause of death in 10 patients (1.4%). Age >60 years, leukocytes >10x109/L, RAS, male gender and serum creatinine level >1.4 mg/dl at presentation were independent prognostic factors for induction death. Patients developing RAS had a higher cumulative incidence of relapse (CIR) in the LPA96 trial (40% vs 15%, p<0.01), but there were no significant differences in the LPA99 trial (12% vs 13%). In conclusion, we have observed a bimodal peak incidence of RAS during the induction phase, with the first peak from days 0 to 7 and the second peak from days 15 to 30. Patients with leukocytes >5x109/L and serum creatinine level >1.4 mg/dl are at high risk for development of RAS, which is an adverse prognostic factor for induction death. The negative impact of RAS on CIR among patients treated with the LPA96 trial was not observed in the LPA99, in which patients received additional doses of ATRA for consolidation therapy.
Retinoic acid syndrome: a problem of the past?
