Acute Promyelocytic Leukemia (APL) prognosis has largely improved over the past decades and is now the acute leukemia with the highest cure rates. However, its early death rate, which reaches 30% in several real-world studies, is the main obstacle to achieving lifelong remission in most patients.

Several reports have suggested a potential role of the Disseminated Intravascular Coagulation (DIC) Score of the International Society of Thrombosis and Hemostasis as a marker of bleeding and death in APL.

Since the APL-induced coagulopathy and hyperfibrinolysis are responsible for most early deaths, a method to assess the severity of the coagulopathy could guide the daily patient management and the intensity of the blood products used.

We aimed to test the hypothesis of whether the improvement of the DIC Score through coagulopathy support can improve the rate of early death in APL.


This retrospective, single-center study, included consecutively admitted patients with newly diagnosed genetically confirmed Acute Promyelocytic Leukemia between 2000 and 2020, who were treated with ATRA + anthracycline protocols and prednisolone 0.5 mg/kg differentiation syndrome prophylaxis. All patients received platelets, fibrinogen, and/or fresh frozen plasma according to published clinical indications.

Baseline demographics and clinical data were collected, and the DIC Score was calculated at diagnosis and after 48 hours for each patient, using fibrinogen, platelets, prothrombin time and d-dimers. "DIC Score improvement" was defined as an increase of ≥1 point in the calculated score at 48 hours; a stable or decreasing score were considered "no-improvement".

A 30-day follow-up for each patient was conducted, and any death during this period was considered in the early-death group.

Statistical analysis was performed using Stata, with chi-square test and Mann-Whitney test for differences between groups, and logistic regression for the analysis of predictors of early death.


A total of 67 patients were included, after excluding 9 patients for missing coagulation values, 8 for first-line arsenic trioxide treatment, and 7 patients for death within 48 hours of hospital admission.

The median age was 53 years (25-82), and 29.9% of patients were aged >60 years old. The 30-day mortality rate for this cohort was 31.3%.

Comparing the baseline differences between the two groups, the only statistically significant difference detected was a less frequent improvement of the DIC Score over 48 hours in the early-death group (42.86% versus 76.09% in the remaining patients, p=0.008). There was no difference in the Sanz Risk distribution nor age between the two groups.

The univariate analysis of predictors of death in 30 days (logistic regression) yielded age >60 years and 48h-improvement of the DIC Score as potential predictors of death. A multivariate analysis model was built incorporating these 2 variables with the Sanz Risk (the number of events in this study permitted only a 3-variable prognostic model), and it identified age>60 years (OR 4.84 [1.37-17.10]; p=0.014) and the lack of a 48-hour improvement in the DIC Score (OR 5.46 [1.61-18.47]; p=0.006) as independent predictors of early death.

In patients under 60 years old, having no improvement in the DIC Score over 48 hours is still associated with almost five times superior odds of 30-day mortality (OR 4.55; p=0.038).


In this cohort, we identified a lack of DIC Score improvement and age>60 years as independent predictive factors of early death in APL.

These findings identified a new independent predictor of early death using a dynamic DIC Score assessment, which is easily accessible and calculated, and illustrate how this score can be applied daily in the clinical care of APL patients.

Together with the current APL treatment strategies, the intensification of measures to control the leukemia-associated coagulopathy in patients who do not show a DIC Score improvement in the first 48 hours after diagnosis could decrease the early death rate in this otherwise highly-curable leukemia.


Brioso Infante:Astellas: Speakers Bureau; Novartis: Speakers Bureau.

Author notes


Asterisk with author names denotes non-ASH members.