Hydroxyurea in sickle cell patients: A look at the correlation with hematologic malignancies

Background: Hydroxyurea (HU) is a chemotherapeutic agent used in sickle cell patients (SCP) to reduce the frequency and intensity of sickle pain crises. It is a myelosuppressive agent with proposed long-term carcinogenic and/or leukemic effects, particularly in myeloid malignancies. Current data has failed to conclusively demonstrate this outcome in its long-term use. Yet many SCP are maintained on HU for years and therefore, could provide evidence of the long-term neoplastic effects of HU. Methods: In a retrospective review of the Wake Forest Baptist Medical Center (WFBMC) Electronic Data Warehouse (EDW) capturing data from 1980 until present, we compared the prevalence of hematologic malignancies in SCP to the general population (GPop) using the appropriate ICD-9 codes (ICD-9 200-208.xx, 238.4, 238.71, 289.83, 277.30). We then examined the use of HU in the SCP versus the GPop to compare the difference in disease prevalence as it related to HU. Finally, the statistical significance was measured with the Fisher’s exact test. Results: Of the 3,874 SCP identified in the EDW, 163/3,874 (4.2%) were on HU. 153/3,874 (3.9%) had a diagnosis of a hematologic malignancy. In the GPop, 222.000 (1.04%) were diagnosed with a hematologic malignancy and 222/2,123,301 (0.01%) were on HU. 46 (1.19%) SCP were diagnosed with a lymphoid or histiocytic neoplasm versus 8,055 (0.37%) in the GPop; 4 (10.87%) SCP were on HU. 23 (0.59%) SCP versus 3,753 (0.18%) GPop had a lymphoid leukemia diagnosis; 4 (17.39%) SCP were on HU. 39 (1.01%) SCP had a myeloid leukemia diagnosis (4,957 (0.23%) GPop); 1 (0.03%) SCP was on HU (p < 0.5884). Conclusions: In a review of data collected over 35 years, there seems to be an increased prevalence of hematologic malignancies in the SCP population, predominantly of lymphoid origin; the use of HU seems to further influence this finding. In contrast, while there is a notable mild increase in myeloid leukemia diagnoses compared to the general population, there does not seem to be the same correlation with HU. Abstract source