INTRODUCTION: Bone marrow examination (BME) is performed in Human Immunodeficiency Virus-infected (HIV+) children with haematologic abnormalities to exclude specific disease (SD). AIMS: To describe the: (1) indications for BME , (2) utility of BME to diagnose SD, (3) patient characteristics associated with SD or non-specific disease (NSD). METHODS: Design: Retrospective review. Definitions: SD: BME positive for opportunistic infection (OI) or HIV-related malignancy. NSD: HIV-related changes only. RESULTS: Eighty six BME’s were done. Suspected SD in 56/86(65.1%) was the most common clinical indication. Bicytopaenia(n=32) and isolated cytopaenia(n=31) were the most common haematologic indications. NSD 48/86 (55.8%) was a more common finding than SD 32/86 (37.2%). Granulomas, pure red cell aplasia and malignancy were the SD identified. Pre- highly active antiretroviral therapy (HAART), advanced stage, and not being virally suppressed were significantly associated with NSD. CONCLUSION: The yield of SD (37.2%) on BME is comparable to adult studies. HAART should be instituted before BME as NSD will be the most likely finding.

Fluorine-18 Fluorodeoxyglucose positron emission tomography-computed tomography (F-18 FDG PET-CT) is now established in the staging, restaging and therapy response monitoring of Hodgkin’s lymphoma (HL) and high grade Non-Hodgkin’s lymphoma (HG NHL), specifically for nodal disease and extra-nodal disease excluding the bone marrow. The role of FDG PET-CT for evaluating bone marrow involvement in HL and HG NHL has not been established yet. There are however several publications on this subject but no consensus has been reached. Bone marrow trephine biopsy (BMB) is the gold standard for bone marrow assessment in lymphoma. Although the occurrence of adverse effects is uncommon, BMB is an invasive procedure that may induce anxiety in patients. A retrospective review of FDG PET-CT bone marrow findings of HL patients referred for a staging scan from June 2008 to January 2014 was done, these findings were compared to the BMB findings also done as part of initial staging. The findings of 55 patients were reviewed analysed.

Cytogenetic evaluation of human bone marrow cells is one of the principal sources of diagnostic and prognostic information in the evaluation of the myeloid leukemias. In the majority of cases, these diseases are characterized by non-random chromosomal changes in the cells of the malignant clone. The chromosomal abnormalities are present only in the leukemic cells, which are distributed along with normal cells in the bone marrow and throughout the circulation. The objective of this thesis was to test the hypothesis that suitable criteria could be established for automated metaphase detection using human bone marrow preparations. This involved computerized, low resolution scanning of a specimen slide, and the measurement of object features which allowed metaphases to be adequately distinguished from nuclei and debris. Two approaches were investigated. The first used a line-scanning system, in which microscope slides were scanned line by line with a linear CCD detector, and focussing was performed automatically. Eighteen signal features were measured for each detected object. Three group discriminant function analysis was performed on objects from a large number of slides from both types of preparations, in order to distinguish metaphases from nuclei and debris. The second method evaluated the use of a frame scanning system. Objects were detected in a frame-by-frame scan of microscope slides, using a two dimensional CD camera. Feature measurements were performed for all objects within a specified area range, and three group discriminant function analysis was performed on data from a large number of slides. In both approaches, the performance of the discriminant functions was evaluated on independent samples collected from a number of patients, in order to determine the operational error rates of the systems. The sensitivity of the line scan system for metaphase detection was 86%, compared to 92% fror the frame scannning system, while the specificity was 84% for the line scan system, and 86% for the frame scanning system. The frame scan system was shown to be useful for determining the mitotic index of cells cultured for varying periods of time prior to fixation. Four patients with AML were examined, and the results of the analysis show that the mitotic indices could be determined in this way to an accuracy of approximately 5%. The mitotic indices differed as a function of time for different patients. / Science, Faculty of / Physics and Astronomy, Department of / Graduate