Central Nervous System Leukemia in Acute Myeloid Leukemia

Introduction: Central nervous system leukemia (CNSL) is less common in patients with acute myeloid leukemia (AML) compared to patients with acute lymphoblastic leukemia. CNSL in AML patients is rare with an incidence less than 5%. Before high dose cytarabine was incorporated into induction and consolidation regimens the incidence was higher, approximately 20%.  Because of the low incidence of CNSL in AML routine evaluation (lumbar puncture with cerebrospinal fluid analysis) in patients with no CNS symptoms is not indicated at diagnosis or relapse.  Patients with acute myelomonocytic leukemia, high risk or relapsed acute promyelocytic leukemia or AML with inversion 16 or chromosome 11 abnormalities may have a higher risk of CNSL. High white blood cell count at diagnosis, elevated LDH and age less than 2 years of age may also increase risk of CNSL. This study was undertaken to evaluate the incidence of CNSL in patients with newly diagnosed and relapsed AML. Methods: 298 adult patients with AML were identified through the Mayo Clinic Arizona Cancer Registry between January 2002 and February 2012. 270 had records available for retrospective review. Data on AML status (newly diagnosed/relapsed disease), cytogenetics, molecular markers, white blood count at diagnosis and lumbar puncture results was collected. Results:9 (4%) patients underwent lumbar puncture. 8 patients had CNS symptoms. CNS evaluation in the asymptomatic patient was performed  because of high white blood cell count and myelomonocytic AML at diagnosis. 6 (67%) of the patients who underwent lumbar puncture were found to have CNSL. Of the patients with CNSL all were symptomatic, 4 had relapsed or refractory AML (1 post allogeneic hematopoietic stem cell transplant (HCT)) and 2 had initial diagnosis of AML. All patients with CNS leukemia had myelomonocytic AML. No trend in cytogenetic subtype was noted. Most patients did not have white blood count at diagnosis or molecular markers available for evaluation.  Discussion: AML can be a difficult disease to cure. Patients with CNSL have worse outcome and need altered treatment regimens. CNSL occurring after HCT has an even more dismal prognosis. Currently there is no indication to evaluate for CNSL in asymptomatic patients either at diagnosis, relapse or pre HCT. This small review is notable because all patients with CNSL had myelomonocytic AML. It has been suggested previously that these patients are at higher risk for CNSL. Perhaps the risk for CNSL in those with myelomonocytic AML has a higher incidence than previously reported and this patient population should have routine evaluation of the CNS. Should asymptomatic patients with myelomonocytic AML undergo lumbar puncture at initial diagnosis, relapse or pre HCT? Would prophylactic administration of intrathecal chemotherapy benefit this population? Further prospective study of the incidence of CNSL in patients with AML would be beneficial.