Differing age patterns in chronic myeloid leukemia by racial ethnic group.

Abstract

1577 Background: The Glivec International Patient Assistance Program (GIPAP) was initiated in 2001 to supply doses of imatinib free of cost to CML patients in developing countries. In reviewing the pattern of CML among the 32,032 patients treated from 90 countries, it was apparent that the age of diagnosis was occurring at a younger age (median: 38 years [y] with 95% < 64y) compared to that of the United States (US) (median: 66y). A descriptive epidemiological study was warranted to describe the age patterns of CML. Methods: Data from the Surveillance, Epidemiology, and End Results (SEER) 17 Registries (1973-2006) and the 72 registries published in Cancer Incidence in Five Continents (CI5) (1958-1997) were analyzed. Results: 18,923 cases of CML were reported in SEER with 25% < 49y. 57% were male and 84% were white (9% black, 7% other). Age frequency distributions at diagnosis showed differences based by race; whites exhibited a peak frequency near 75y, whereas blacks had a bimodal age distribution (40 y and 70 y). Other races exhibit a shift in the age distribution to younger patients. The age-adjusted rate per 100,000 was 1.67 with a higher rate in males compared to females (2.19 vs. 1.29). Incidence rate curves and age-period-cohort (APC) fitted age-at- onset curves were similar for males versus females. Rates were similar for white and black (1.65 vs 1.68). However, an age-specific incidence rate crossover was observed between black and white cases with rates higher among blacks < 50 y and rates higher among whites > 70 y. Data from CI5 show a bimodal distribution (35-44 y and 60-74 y). Conclusions: Bimodal age distributions at diagnosis by race and ethnicity have emerged from this analysis of three different datasets. The GIPAP cohort suggested a younger age at diagnosis. The differential age pattern was supported by the SEER analysis suggesting an earlier age at onset (30-40 y) and a later age of onset (65-75 y). Crossover in incidence rates between black and white was consistent. A similar bimodal distribution among CI5 was also concluded. To our knowledge, this is the first report of bimodal CML, suggesting two distinct populations in an otherwise homogeneous disease. Additional analytical studies are needed to determine the source of this heterogeneity. No significant financial relationships to disclose.