Characterization and analysis of the outcome of adults with acute myeloid leukemia treated in a Brazilian University hospital over three decades

We evaluated the outcome of 227 patients with acute myeloid leukemia during three decades (period 1 1980’s, N = 89; period 2 1990’s, N = 73; period 3 2000’s, N = 65) at a single institution. Major differences between the three groups included a higher median age, rates of multilineage dysplasia and co-morbidities, and a lower rate of clinical manifestations of advanced leukemia in recent years. The proportion of patients who received induction remission chemotherapy was 66, 75, and 85% for periods 1, 2, and 3, respectively (P = 0.04). The median survival was 40, 77, and 112 days, and the 5-year overall survival was 7, 13, and 22%, respectively (P = 0.01). The median disease-free survival was 266, 278, and 386 days (P = 0.049). Survival expectation for patients with acute myeloid leukemia has substantially improved during this 30-year period, due to a combination of lower tumor burden and a more efficient use of chemotherapy and supportive care.

Fifteen years ago, our group reported the treatment and outcome of AML at our institution, and discussed several reasons for the low remission and survival rates observed, including poor hospital convenience, health system deficiency, patients' poor socioeconomic background, and different biological aspects of AML.A remarkable finding was that ~40% of patients did not receive any chemotherapy because of very poor general conditions, and this was the main reason for the poor results achieved (14).
In order to confirm recent improvements in treatment outcomes, we retrospectively reviewed the clinical characteristics, treatment and outcome of all patients with a diagnosis of AML admitted to Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, over the past 30 years.

Material and Methods
This retrospective cohort study included all patients diagnosed with AML at Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Brazil, from January 1979 to December 2008.Patients were identified from hospital medical records and the Hematology Service registry.Patients with acute promyelocytic leukemia (APL) and those previously treated at other institutions were excluded.The study was approved by the Hospital Universitário Clementino Fraga Filho Ethics Committee.
The diagnosis of AML was based on procedures available at the time, including bone marrow aspiration and biopsy, and cytogenetic and immunophenotype analyses.Cases were classified according to French-American-British www.bjournal.com.brBraz J Med Biol Res 44( 7) 2011 (FAB) criteria (15).For the purpose of this study we reviewed the slides for FAB classification and for the detection of the presence of multilineage dysplasia.The treatment approach varied between time periods; for the purpose of the present analysis, treatment was classified as curativeintent (when standard induction was given), palliative-intent (other regimens), and no chemotherapy.Standard remission induction was based on treatment with cytarabine plus an anthracycline in most patients.Until 1985, the regimen consisted of 100 mg/m 2 thioguanine administered orally every 12 h on days 1 to 7, 100 mg•(m 2 ) - Complete remission (CR), treatment failure and relapse were defined according to the International Working Group for AML (16) and the recommendations from an international expert panel (17).Briefly, CR required <5% blasts to be present in a bone marrow aspirate sample, with an absolute neutrophil count >1000/μL and a platelet count >100,000/μL.
Overall survival (OS) was calculated from the diagnosis of AML until the date of death or last follow-up.Disease-free survival (DFS) applies only to patients who obtained CR, and was calculated from the time of CR assessment to the date of relapse, last follow-up or death.
Dichotomous variables were compared by the Fisher exact test or the chi-square test, and continuous variables were compared by the Kruskal-Wallis test.Actuarial curves of OS and DFS were constructed using the Kaplan-Meier method, and compared by the log-rank test.Predictors of outcome (OS) were assessed by Cox regression analysis, and the results are reported as hazard ratio (HR) with the corresponding 95% confidence interval (95%CI).Variables with P value <0.1 by univariate analysis were entered in multivariate analysis.P values were two-sided, and were considered to be statistically significant with values <0.05.All statistical analyses were performed using the SPSS 11.0 software (SPSS Inc., USA, 1989USA, -2001)).

Results
Of 295 patients with AML identified in hospital records, 68 were excluded for the following reasons: 52 with APL, 11 with biphenotypic leukemia, and 3 previously treated at other institutions.Among the remaining 227 patients, 125 were males.The median age was 45 years (range 12-91), and 25% were >60 years old.The most common FAB subtypes were M4 and M2 (27 and 22.5%, respectively).AML with multilineage dysplasia was present in 18.5% of patients.
Fifty-eight patients (25.5%) did not receive treatment.Among the 169 patients receiving remission induction chemotherapy, 75% were treated with cytarabine in continuous infusion for 7 days plus an anthracycline for 3 days (7+3 regimen), 15% received TAD and 10% received other regimens.Fifteen patients died before day 14 of remission induction and were not assessed for response to chemotherapy.Among the remaining 154, CR was achieved in 88 (57.1% of patients receiving remission induction), after a single course of induction in 72, and after a second course in 16 (5+2 regimen).The median time (days) to achieve >1000 neutrophils/μL and >100,000 platelets/μL was 27 (range 4-153) and 28 (range 9-104) days, respectively.

Comparative analysis of three decades
The number of cases for each time period and their characteristics are detailed in Table 1.Major differences in baseline characteristics included a higher proportion of multilineage dysplasia and higher median age over time www.bjournal.com.brBraz J Med Biol Res 44( 7) 2011 (and, as a consequence, a higher rate of co-morbidities in recent years).By contrast, in recent years the rate of clinical manifestations of advanced leukemia, such as hepatosplenomegaly, bleeding, fever, documented infection, and baseline leukocyte count was lower.

Discussion
Our study shows that some characteristics of adult patients with AML have changed over the past 3 decades, with a higher proportion of older patients, with co-morbidities and multilineage dysplasia.The higher proportion of elderly patients is expected as a consequence of the increasing longevity of the population (18).Indeed, in Rio de Janeiro, during the same period there was an increase of people more than 60 years old, from 8.7% in the 1980's to 11.2% in the 1990's, and 12.8% in the 2000's (19).Furthermore, AML in the elderly is associated with several biological and clinical features, including a high frequency of co-morbidities and multilineage dysplasia, as observed in our cohort (20).
In addition, despite the higher incidence of these characteristics, the outcome improved over time.This could be a reflection of a lower disease burden, as observed by the lower frequency of lymphadenopathy, hepatomegaly, splenomegaly, fever, bleeding, and infection at diagnosis, as well as the median leukocyte count at baseline.In a previous report from our group, 37% of patients with AML were too sick at diagnosis to receive any treatment (14).
In the present study, we observed that the proportion of patients receiving treatment increased ~10% per decade: 66% in the 1980's, 75% in the 1990's, and 85% in the 2000's.Taken together, the increased proportion of patients being treated and the lower tumor burden on admission suggest that the primary health care system of Brazil has generally improved.
Moreover, the 5-year overall survival may have improved not simply as a consequence of a higher proportion of patients receiving chemotherapy, since there was a trend for an increased OS among treated patients (from 10% in the  1980's to 26% in the 2000's).The most significant improvement in the outcome can be appreciated in the 5-year DFS curve.In the 2000's, DFS was three times higher than that of the 1980's (44 vs 13%).This is probably a result of better consolidation or intensification employed more recently, including HCT, as well as progresses in supportive care, since there were no significant differences in the rates of complete remission (42.4% in the 1980's, 60% in the 1990's, and 54.5% in the 2000's).Indeed, when we excluded patients undergoing HCT, the differences in OS and DFS were much lower, and lacked statistical significance.
The multivariate analysis of prognostic factors showed that decade was significantly associated with the outcome.This, most probably, reflects a combination of elements related to diagnosis and treatment: effective induction treatment, higher proportion of treated patients, intensive postremission therapy, better management of chemotherapyinduced toxicity, appropriate use of antimicrobial agents, development of autologous and allogeneic transplantation, and progresses in overall supportive care (4,9,18,21).
Among the other variables, age has proved to be an important prognostic factor in many studies (4,(22)(23)(24)(25).In addition, AML with multilineage dysplasia has been associated with a poor prognosis.In our analysis, this variable was not significant by multivariate analysis, probably due to a strong correlation between multilineage dysplasia and older age (26).
Other prognostic factors identified in our analysis have been reported in other studies as important predictors of outcome, including kidney disease (23,27,28), hepatomegaly (29), and bleeding (30).The presence of the FAB M2 subgroup was protective, possibly as a consequence of the relationship between FAB M2 subgroup and well-defined prognostic factors such as t(8;21) translocation (31), and a lower frequency of FLT3 mutations (32).
The present study suffers from the limitations of retrospective studies, especially the lack and inconsistency of data collected from patients' charts, without a prospectively collected case report form.In addition, the analysis of prognostic factors is hampered by the fact that cytogenetics, an important prognostic factor in AML (33,34), was performed in a minority of patients from this cohort.

Figure 1 .
Figure 1.Overall survival as a function of decade of diagnosis.A, All patients; B, treated patients.

Figure 2 .
Figure 2. Disease-free survival as a function of decade of diagnosis.

Table 1 .
Characteristics of 227 patients with acute myeloid leukemia stratified by decade of diagnosis.