Clinical and biological features of acute promyelocytic leukemia patients developing retinoic acid syndrome during induction treatment with all-trans retinoic acid and idarubicin

Although all-trans retinoic acid (ATRA) is generally well tolerated, some patients develop a potentially severe and life-threatening complication referred to as retinoic acid syndrome (RAS). We analyzed here the biological and clinical characteristics of 110 consecutive patients with genetically proven APL, with the aim of identifying predictive features of developing RAS.

All patients described were treated with AIDA1 or AIDA 20002 protocols, between January 1993 and December 2005 at the University “La Sapienza” of Rome. Diagnosis was confirmed at the genetic level by RT-PCR identification of the PML/RARa hybrid as previously described.1 The presence of FLT3 internal tandem duplication (ITD) was investigated in 80 cases using the technique previously reported.3 Immunophenotype was performed by flow-cytometry using a wide panel of monoclonal antibodies including CD13, CD33, HLA-DR, CD34, CD2, CD7, CD15, CD9, CD117, CD56, MPO (Becton Dickinson, Mountain Flow, CA, USA), considering a sample antigen-positive if >20% of the cells reacted with a specific monoclonal antibody, whereas for CD34 a cut-off of >10% was used. According to Frankel et al.4, the diagnosis of definitely present RAS was clinically established by the presence of at least three of the following signs: weight gain, respiratory distress, unexplained fever, interstitial pulmonary infiltrates, pleural or pericardial effusions. For statistical analysis, the Wilcoxon-Mann-Whitney test was performed for comparison of non-parametric series and Fisher’s exact test was used to compare categories. Values of p<0.05 were considered as statistically significant. Overall survival was measured from the time of diagnosis to death or last follow-up.

Table 1.Clinical manifestations and outcome of patients developing retinoic acid syndrome.

Median time elapsed between treatment initiation and first symptoms of RAS was four days (range 3–5). At the time of initial suspect of RAS, the median WBC count was 2.9×109/L; in 5/15 patients RAS was accompanied by an increase of WBC up to 10×109/L. Median peak WBC was 5.3×109/L and median doubling time was four days. Respiratory distress was the first manifestation of RAS in 12/15 patients (80%, Table 1). CT scan was performed at the earliest manifestation of respiratory distress in all patients. Pulmonary infiltrates were documented in 14/15 patients and in 3 of them without dyspnea. Fever was present in 12 patients, weight gain in all, renal failure in 2, pleural/pericardial effusions were present in 3 patients.

Treatment of RAS consisted of dexamethasone 10 mg/m2/bid until complete disappearance of symptoms, administered for a median of six days. ATRA was discontinued only in 2 cases in which RAS was considered life-threatening and did not improve with dexamethasone.

Resolution of RAS occurred in a median time of four days (range 2–8 days) from dexamethasone initiation. All patients achieved complete hematologic remission after a median of 30 days (range 28–45), were given 3 cycles of consolidation and obtained molecular remission at the end of consolidation. Five patients (33%) underwent disease relapsed at a median time of 22 months (range 12–32). Four patients died in second relapse whereas one patient is alive after allogeneic stem cell transplantation after 72 months. Presenting features and treatment outcome of the 15 patients who developed RAS and of the 95 who did not are shown in Table 2. Significant differences were observed in the prevalence of M3v FAB subtype (50% vs. 26%, p=0.02), median WBC count (6×109/L vs. 2.8×109/L, p=0.01), prevalence of high relapse risk (46% vs. 24%, p=0.001), bcr3 PML/RARα (66% vs. 44%, p=0.001), FLT3-ITD (54% vs. 36%, p=0.002), expression of CD2 (54% vs. 12%, p=0.0001) and CD15 which was only detected in patients developing RAS (p=0.0001). No differences in CR rate and in overall survival were observed between the two groups (40 vs. 45 months, p=0.35). Finally, patients who experienced RAS had a 33% rate of morphological relapse as compared to 11% of patients who did not (p=0.002).

Table 2.Comparison of clinical/biological features in patients with and without retinoic acid syndrome.

We report here an incidence of RAS of 13.6% from which none of the patients died. The difference in the reported RAS frequency and mortality among different series may reflect several variables including its definition and recognition criteria, type of concomitant therapeutic protocol for APL and RAS treatment itself.5,6 To better define biological features associated with bona fide RAS, we included in the present study only patients in which a definitely present RAS could be diagnosed according to the criteria of Frankel et al.4 As to the absence of deaths due to RAS in our series, this may reflect improvements in early recognition of the syndrome and prompt institution of dexamethasone. As reported by Vadhat et al.7, we found a significant correlation with WBC count at presentation; however, we did not find that WBC count above 5×109/L on day 1, or above 6×109/L, 10×109/L, 15×109/L, on days 5, 10 or 15 of ATRA treatment respectively, was predictive for the development of the syndrome as observed by Fenaux et al.8 We documented an aberrant phenotype in all patients who developed RAS, in particular a strong associations with CD2 and CD15 expression. CD2 can mediate the adhesion to several molecules such as the ubiquitously expressed CD58 or CD59 expressed on platelets and erythrocytes, and may play a role in leukoagglutination, contributing to tissue damage by microvascular occlusion.9 As to CD15, it is known that this antigen mediates the adhesion to activated endothelial cells through selectins ligand, and favors the first rolling event.10,11 Although in the present study we did not evaluate the in vivo induction of CD2 and CD15 levels during treatment, others have reported that both CD2 and CD15 expression are modulated by ATRA suggesting a role for these antigens in the development of RAS.11

The frequent association of the FLT3-ITD mutation in APL blasts at presentation and its correlation with elevated WBC count and the PML/RARa bcr3 isoform is well known.12 Recently, Marasca et al.12 reported gene expression profiling and FLT3 mutational status in a series of APL patients, and found that patients ITD had increased expression of genes regulating blood coagulation (CD97, PTX3, H963) and cell adhesion (AMIGO2, LGALS 1–2) suggesting a role for FLT3-ITD in the development of RAS during induction.

To summarize, our study suggests that the risk of developing RAS in APL patients is associated with consistent phenotypic and genotypic features of leukemic cells. Further studies are warranted to investigate the mechanistic link between these features and the pathogenesis of RAS in these patients.