Abstract
All-trans retinoic acid (ATRA) has markedly improved the outcome of APL: however, the incidence of thrombotic events seems to be higher in the ATRA treated patients. To assess the true incidence of such complications, we compared 2 different cohorts of APL patients: 37 patients (M/F 17/20, median age 37 years, range 14 – 58) treated with chemotherapy (CHT) alone (GIMEMA 0389-group A) and 90 patients (M/F 40/50, median age 43.5 years, range 19 – 75) treated with ATRA + CHT (AIDA and AIDA 2000 protocols-group B). In the group A, no patient had thrombotic complications during induction or consolidation treatment: however, among 15 patients who relapsed and received a reinduction treatment with ATRA +/− CHT, 3 (20%) developed a thrombotic disease [1 acute myocardial infarction (AMI), 1 pulmunary embolism, 1 deep venous thrombosis (DVT)]. In the group B, 11/90 patients (M/F 5/6, median age 60 years, range 32 – 71) (12%) had a thrombotic event. Of them, 8 patients developed it during induction (4 episodes of unstable angina, 2 right ventricular thrombosis, 2 DVT), after a median time of 26 days from ATRA treatment (range 3 – 78), a median PLTS count of 208 x 109/l (3/8 patients showed a PLTS count < 50 x 109/l when the thrombotic event occurred) and without prothrombotic biochemical abnormalities. The remaining 3 patients developed a DVT during consolidation phase, after 31, 34 and 37 days respectively from start of 1st consolidation cycle and with a normal PLTS count. No association was observed with ATRA syndrome or haemorrhagic diathesis at presentation, whereas 8/11 patients showed the presence of FLT3-ITD at diagnosis. Phenotypic characterization revealed a classic panel in all patients (CD13+, CD33+, CD9+, MPO+), but a peculiar positivity for CD2 was seen in 7 patients, for CD15 in 3 patients and for both antigens in 1 patient. CD2 antigen is normally found on T cells and mediates adhesion to CD58: it is possible that this aberrant expression on the surface of APL cells play a role in the leuko-agglutination during ATRA treatment, contributing to thromboses in peculiar site, such as right ventriculum. In the 3 patients CD2 negative, we found a similar aberrant expression of CD15, that has been reported in literature as an antigen normally modulated by ATRA, together with other adhesion molecules. The association between CD2 expression, short type of bcr transcript and FLT3 abnormalities has been also confirmed. Our data outline the increase of thrombotic risk linked to ATRA treatment in APL patients, with an incidence of 10 – 15% to be confirmed in other series. We suggest a possible and crucial role of aberrant features at diagnosis, such as phenotype and molecular abnormalities, to explain this high thrombotic incidence: thus, further prospective studies are warranted to analyse clinical and biological factors predisposing to thrombotic disease in such patients.
Author notes
Corresponding author