Administration of All-Trans Retinoic Acid (ATRA) to Newly Diagnosed Patients (pts) with Acute Promyelocytic Leukemia (APL) Is Delayed Even At Experienced Centers and Associated with An Increased Early Death Rate (EDR): A Retrospective Analysis of 205 Pts

Early death in APL, most often due to bleeding, has emerged as the major cause of treatment failure now that curative strategies exist. Despite the routine use of ATRA, EDR remains high (Lehmann et al Leukemia 2011, Park et al Blood 2011). Although the optimal strategy to prevent early death is not clear, the recommendation is to initiate ATRA immediately at first suspicion of the disease without waiting for genetic confirmation. Therefore, we examined the timing of ATRA administration.

To determine time interval from initial presentation to ATRA administration, we retrospectively collected data on all newly diagnosed APL pts presenting between 1992–2009 to 4 institutions: Northwestern University, Chicago, IL (university medical center); Memorial Sloan-Kettering Cancer Center, New York, NY (free standing cancer center); John J. Stroger Hospital of Cook County, Chicago, IL (public hospital); and Rambam Medical Center, Haifa, Israel, (Northern Israel's largest hospital; a tertiary referral center). We also examined 3 other time intervals: presentation to suspicion of APL, suspicion of APL to ordering ATRA, and ordering ATRA to its administration.

We identified 205 newly diagnosed APL pts: 46% men, median 48 years (range 1.5–85). Median white blood cell (WBC) count at presentation was 2,100/uL (range 300/uL-222,500/uL); 25% had high risk (HR) disease (WBC >10,000/uL). Median time interval from initial presentation to suspicion of APL was 1 day and median time from suspicion of APL to ordering ATRA was an additional day (table 1). ATRA was ordered on the day APL was suspected in 32% pts, the next day in 31%; 2 days after suspicion in 17%; and after 3 or more days in 16%. 89% received ATRA on the day ordered. At least 1 bleeding episode was identified in 34% of 185 pts with bleeding data available. Bleeding was associated with higher WBC count (p=0.0003) and lower hemoglobin (p=0.027) at presentation. 46 of 186 pts with complete information on time from presentation to administration of ATRA died. 23 (12%) died within 30 days of presentation; comprising half of all fatalities. Causes of death were: hemorrhage −15, sepsis −4, suspected MI −2, pneumonia −1, and sepsis plus differentiation syndrome -1. Among deaths within 30 days, 48%, 22%, 26% and 7% were in 1st through 4th weeks, respectively. 4 (18%) of these 23 pts died before ATRA was administered, all day 1 or 2 after presentation and all from bleeding. Only 15/182 pts received ATRA on day of presentation. Two of these 15 (13%) died within 30 days (none from bleeding). In comparison, 7/40 (18%) who received ATRA on day after presentation died within 30 days (71% from bleeding). 10/127 (8%) who received ATRA on day 2 or after died within 30 days (6 from bleeding). 20% in each group who received ATRA on either day of presentation or day 1 after presentation had HR disease. For this subgroup, if ATRA was administered on the day APL was suspected or one of the following 2 days, EDR was 19% (7/37). However, if ATRA was initiated on day 3 or 4, EDR was 80% (4/5); (p=0.013). 59% received ATRA prior to confirmation and 41% received ATRA on the day APL was confirmed or later.

(1) APL was suspected rapidly upon presentation, usually within 1 day and ATRA was almost always administered on the day ordered. (2) However, ATRA was not given to most pts the day APL was suspected and for some 2 or more days elapsed. This time interval contributed to the delay in ATRA administration, suggesting physicians waited for marrow morphology or genetic confirmation before ordering ATRA. (3) Although at these centers a lower EDR (12%) than reported from the SEER database (17.3%) was observed, the current recommendation to give ATRA at first suspicion of APL was often not practiced. (4) There appears to be an association between EDR and timing of ATRA administration; other factors contribute to the EDR including variability in blood product support. (5) Our study argues that educating health care professionals who are the first to encounter APL pts as to the urgency of ATRA administration will reduce early deaths that occur even in the ATRA era.

No relevant conflicts of interest to declare.