Chemotherapy for Acute Myeloid Leukemia Can Compensate Other Adverse Prognostic Factors in Elderly: A Retrospective Longitudinal Cohort Study in a Local Community, Japan

Japan is one of the prominent aging countries, in where incidence and mortality of malignancy in elderly is increasing which is recognized as a concerned national health care issue. Among various types of cancer, incidence and prevalence of elderly acute myeloid leukemia (AML) is increasing as well, and in the most cases those patients have concomitant diseases. Thus far, the treatment for AML in elderly has been generally realized to reduce the dose intensity due to impaired organ function and comorbidity. However, the utility of dose reduction in elderly AML has not been well understood. Thereby, we performed a retrospective population cohort study of AML in a local community in Japan in order to investigate practical risk factor based on general patient population including more than 65 to 75 years old.

We reviewed medical charts and medical records of AML patients diagnosed at seven institutions in Kagawa prefecture between January 1, 2006 and December 31, 2010. We collected patients' characteristics, background, and clinical information including laboratory data, adverse events and outcome. Induction therapy regimen was practically decided by each treating physician based on available clinical data and local standards of care, but not on karyotype. We categorized chemotherapy regimens for non-APL patients into 3 groups; intensive chemotherapy, less-intensive chemotherapy, or best supportive care.

A total of 213 patients (127 males and 86 females) were diagnosed with AML. With an average population during study periods in Kagawa, the incidence of AML is 4.26 per 100,000 per year. The male to female ratio was 1.48. The median age was 70 years (average 67, range 24∼95). There were 16 APL patients and 197 non-APL patients in this cohort. The 5-year overall survival (OS) rate was 21.1%. In patients who are likely to meet criteria usually applied in clinical studies (i.e., de novo AML with PS 0 to 2 and no renal dysfunction), the 5-year OS rate was 31.5%. Among 197 non-APL patients median age was 70 years (range 24∼95) consists of 119 males and 78 females. The chromosomal karyotype is known to be the strongest prognostic predictor, and each study group advocates the different classification. We classified our cohort according to 5 different karyotype classifications (NCCN, BSCH, MRC-AML10, SWOG and CALGB). On the whole, 4.4∼5.9% of the patients were classified as having favorable risk karyotypes, and 17.7∼29.1% of the patients had unfavorable risk karyotypes. In terms of treatment intensity, 51.8% patients (102/197) received intensive induction chemotherapy. A majority (71.6 %) of patients ≤ 64 years were treated with intensive chemotherapy, while approximately half (46.0%) of patients aged 65 to 74 years, and only 35.6% of patients ≥ 75 years received intensive chemotherapy as induction therapy (P < 0.001). A complete remission (CR) rate in patients treated was 67.6% (69/102) with intensive regimen, 30.4% (17/56) with less-intensive regimen, and 0.0% (0/39) with best supportive care (P < 0.001). Eight week mortality was 6.9% (7/102), 19.6% (11/56), and 41.0% (16/39), respectively (P < 0.001). In non-APL patients, the 5-year OS rate was 10.6%. Subsequent analysis by age group showed that the 5-year OS rate declined with age; the 5-year OS rate and the median survival of non-APL patients ≤ 64 years, 65∼74 years, and ≥ 75 years were 41.5% and 19 months, 14.1% and 10 months, and 8.9% and 7 months, respectively (P = 0.003). Multivariate analysis revealed aging older than 65 year-old, best supportive care, poor PS more than 3, antecedent hematological disease, and unfavorable risk karyotypes were independent adverse prognostic factors associated with OS.

This analysis provides virtual data from an unselected AML population in a Japanese cohort. Our data can be applied for realistic risk assessment for AML patients including elderly. The present result indicates all prognostic factors has comparable impact on survival, therefore chemotherapy can compensate other adverse prognostic factors even in an elderly patient with poor PS. In aggregates, we conclude it is the best way to adapt as intensive chemotherapy as possible to improve outcomes in the treatment of elderly AML. In further investigation, the external validity of the risk assessment based on identified prognostic predictors from our cohort should be reevaluated for a third party cohort from other countries.

No relevant conflicts of interest to declare.