Clinical features and outcomes of 134 Brazilians with acute promyelocytic leukemia who received ATRA and anthracyclines

Abstract

We report an increased incidence of high relapse risk features in 157 APL Brazilian patients. Out of 134 patients treated with ATRA and anthracyclines, only 91 (67.9%) achieved remission because 43 (32%) died during induction. The death rate during consolidation was 10.5%. Bleeding complications were the most frequent cause of failure (21.6%).

There is sufficient evidence in literature to support the belief that all-trans retinoic acid (ATRA) and concomitant anthracycline-based chemotherapy should be the treatment of choice for newly diagnosed acute promyelocytic leukemia (APL).1,2 In Brazil, APL accounts for more than 20% of acute myelogenous leukemias (AMLs),3 a higher incidence than that reported in developed countries. Nevertheless, despite the fact that anthracyclines and ATRA are widely available, the results with standard treatment are not known.

We retrospectively analyzed medical chart data of 157 APL patients treated from January 2003 to March 2006 at 12 Brazilian institutions. The diagnosis was based on detection of the t(15;17) chromosomal translocation by cytogenetic analysis or of PML/RARα rearrangement by RT-PCR analysis. Laboratory diagnosis of disseminated intravascular coagulation (DIC) was based on changes in activated partial thromboplastin time, prothrombin time, fibrinogen degradation products (FDPs), and/or D-dimers. Central nervous system, pulmonary, or gastrointestinal hemorrhages were considered a severe bleeding. Patients were classified according to the risk of relapse on the basis of WBC and platelet counts (PLT) at diagnosis: low risk, WBC ≤ 10×10/L and PLT > 40×10/L; intermediate risk, WBC ≤ 10×10/L and PLT ≤ 40×10/L; and high risk, WBC > 10×10/L.4

Survival analysis was carried out for 134 patients who received anthracyclines (daunorubicin or idarubicin) plus ATRA in induction, ATRA and anthracyclines (daunorubicin, idarubicin, mitoxantrone or pharmarubicin) with or without citarabine in consolidation, and long term low dose chemotherapy in maintenance as proposed by Fenaux et al.5 Blood bank support was available in all the centers, however, not all of them adopted prophylactic transfusions based on fibrinogen concentrations. Early mortality was defined as death within 14 days of diagnosis. Survival rates were estimated by the Kaplan-Meier method, and compared using the log-rank test. Differences among the risk groups regarding frequencies of bleeding and DIC at diagnosis, and mortality rates were compared using the χ test. APL patients represented 28.2% of AML cases in the analyzed population. This is consistent with the previously reported higher frequency in patients with Latino ancestry.6,7 The median WBC counts (Table 1) was higher than those reported in literature.4,5 Consequently, the frequency of high-risk patients was significantly higher than that reported by PETHEMA and GIMEMA4 (36.9 vs 22.6%, p=0.009). Although the time taken to reach specialized care was not accessed, studies on other hematological malignances suggest that this factor may be associated with the frequency of high tumor burden. The incidence of severe bleeding did not differ from that previously reported8 but was associated with high mortality. DIC, severe bleeding and early mortality were more frequent in the high-risk group (p=0.015, p=0.001 and p<0.001 respectively) (Table 1). PML-RARα isoform distribution did not differ from that described in nonlatino populations.9 This is in contrast to the reported excess of the BCR1 subtype in Mexican Mestizos.10

Table 1.Clinical and laboratory features and causes of mortality.

Among 134 patients available for survival analysis, 13.4% died within 5 days of diagnosis and 26.4% died within the first 14 days. This was mainly due to bleeding (66.6%). Induction and consolidation mortality were 32.0% and 10.5% respectively (Table 1). Three patients (2.2%) relapsed after consolidation and died. All patients alive after induction where in hematologic remission. The overall survival curves of all patients and of those who survived beyond induction are shown in Figures 1A and B respectively. The comparative analysis of these two curves reinforces the hypothesis that support during induction is the major issue to be addressed in developing countries. No significant differences were found between centers (data not shown). As high risk patients had a higher early mortality (Table 1), risk classification may identify patients who need a specific assessment, not only in consolidation, but also more intensive supportive care during induction.

Figure 1.Overall survival of APL patients treated with ATRA in combination with anthracyclines in Brazil. A. Analysis of all patients. B. Analysis excluding patients who died during induction.

Two important issues require particular attention. In the present study, no patient was excluded on the basis of age or performance status. This contrasts with published clinical trials and suggests that the prognosis of APL is not as favorable as is sometimes stated. Neither is the availability of drugs per se sufficient to reduce the gap in outcome between APL patients in developed and those in developing countries. Quicker diagnosis and better supportive care are required. With this aim, the International Consortium on Acute Promyelocytic Leukemia created a unified simplified treatment protocol and a support network offering online bi-weekly conferences with specialists, guidelines and monitoring of supportive care, centralized monitoring of treatment response by molecular biology methods, and an internet data base of patients from Brazil, Mexico and Jordan. We hope that this program will help change the disappointing results reported here.

Acknowledgments

the authors would like to thank Margaret Carbaugh for her critical review of the manuscript