Risk factors in the occurrence and diagnosis of myeloid neoplasms

Thesis (Ph.D.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you.Risk factors for the development and diagnosis of myeloid neoplasms (MN) are not well understood. We conducted three studies using the United Kingdom based General Practice Research Database, a large database with information on medical conditions and drug prescriptions. Benzodiazepines (BZDs) are drugs that interact with a receptor in hematopoietic cells involved in apoptosis, but their effect on MN is unknown. In the first study, we estimated the relative risk of MN in patients taking BZDs. There was a slightly increased risk for myelodysplastic syndromes (MDS) (adjusted odds ratio [OR] 1.6, 95% confidence interval [CI] 1.1, 2.2) and a smaller increased risk for acute myeloid leukemia (AML) and primary myelofibrosis (PMF). However, there was no dose-response relationship, making a causal effect unlikely. Unlike BZDs, certain immunosuppressant drugs are known to increase the risk of MN, although it is unclear whether autoimmune disorders that these drugs are used for also increase the risk of MN. In the second study, we estimated the relative risk of MN in patients with autoimmune disorders, both with and without treatment, as compared with patients without autoimmune disorders. There was a slightly increased risk of MDS in patients with any autoimmune disorder (OR 1.5, 95% CI 1.1, 2.0). Autoimmune disorders also conferred a small potential increase in risk of essential thrombocythemia (ET), PMF, and other myeloproliferative neoplasms. However, it is difficult to determine whether autoimmune disorders increased the risk of MN, or whether the small elevation in the effect estimate was an artifact of misdiagnosis resulting from early MN symptoms that are similar to those of autoimmune disorders. In the third study, we explored the role of surveillance bias in MN diagnosis. There was an increased risk of MDS in patients with a history of frequent blood tests (OR 2.4, 95% CI 1.3, 4.5) that was not present for AML (OR 1.1, 95% CI 0.5, 2.3) or chronic myeloid leukemia (OR 0.7, 95% CI 0.2, 3.1). The increased risk for MDS may be a result of the long-term physician monitoring prior to a diagnosis, or surveillance bias leading to a diagnosis of otherwise indolent cases