Predictive Factors for All-Trans Retinoic Acid Related Differentiation Syndrome In Acute Promyelocytic Leukemia.

All-trans retinoic acid (ATRA) used for the treatment of acute promyelocytic leukemia (APL) can lead to development of differentiation syndrome (DS), a possibly life threatening complication. DS is reported in 2–27% of APL patients treated with ATRA. Since ATRA is metabolized by cytochrome P450 (CYP) enzymes, we sought to identify drug interactions that might be associated with a higher risk for the development of DS in a series of APL patients treated with ATRA. We also evaluated overall incidence of DS, intensive care unit (ICU) admission rates, length of stay (LOS) in hospital as well as other predictive factors related to the incidence of DS.

We identified 60 patients with confirmed APL treated at our institution between May 2004 and January 2010. All patients received ATRA 45mg/m²/day. Patients received concurrent chemotherapy with cytarabine and daunorubicin, idarubicin, arsenic or gemtuzumab as part of induction in 16 (27%), 32 (53%), 10 (17%), and 3 (5%) of cases respectively. Diagnosis of DS was made by the presence of the following symptoms based on the definition by Frankel et al. (Annals of Internal Medicine. 1992; 117: 292–296): dyspnea, unexplained fever, weight gain more than 5kg, unexplained hypotension, acute renal failure, hyperbilirubinemia and chest radiograph demonstrating interstitial infiltrates and pleural or pericardial effusion. Patients with ≥ 4 symptoms of DS were classified as severe DS, while patients with 2–3 symptoms were classified as having moderate DS. Patients were retrospectively evaluated for the development of DS for 5 weeks after initiation of ATRA. Medications were followed in patients until the development of DS. For patients who did not develop DS, medications were followed for 21 days.

Characteristics of the 60 APL patients included: Median age 44 years (18-79 years), 60% female, median white blood cell (WBC) count on admission 1.8 × 10⁹/L (0.4-67.9 × 10⁹/L). Twenty nine patients (48%, 90% confidence interval 37–60%) experienced DS (14 severe and 15 moderate). Dyspnea, unexplained fever and interstitial infiltrates were the most common features of DS. Twenty six of the 29 patients (90%) who met criteria for severe or moderate DS received treatment with dexamethasone. The mean time to development of DS was 9.4 days (range 1–31 days) with a median of 5 days. Seventeen patients developed DS ≤ 8 days after initiation of ATRA; the other 12 patients who developed DS did so between days 13 and 31 after initiation of ATRA. We did not find any difference in overall incidence of DS whether patients were on concurrent CYP 2C8, 2C9 or 3A4 inhibitors, inducers or substrates. Peripheral blood blast counts > 70% versus ≤ 70% on admission was associated with developing DS (p=0.05 using Fisher's exact test). In addition, patients with BMI > 40 kg/m² versus ≤ 40 kg/m² were more likely to experience DS (p=0.00068 using Fisher's exact test). There were no differences in rate of DS in regards to high WBC count, LDH, and albumin on admission. In addition, there were no differences in DS rate in patients receiving different chemotherapy regimens with ATRA. Thirteen (22%) patients were admitted to the ICU (31% in DS patients; 13% in no DS patients; NS). Out of the 29 patients with DS, there were 4 early deaths of which 2 were attributed to DS compared to no early deaths in the patients who did not develop DS (p=0.05). LOS was similar between patients with DS and without DS, 33 days (4-59 days) versus 34 days (19-53 days), respectively.

DS may be underreported in the literature, given that this study is the first to find a high (48%) incidence of DS in APL patients treated with ATRA. Patients with peripheral blast counts >70% or BMI > 40 kg/m² on admission had an increase risk of developing DS. Concurrent use of drugs metabolized in the same pathway as ATRA did not result in increase incidence of DS.

No relevant conflicts of interest to declare.