A Research Report submitted to the Faculty of Health Sciences, University of Witwatersrand, Johannesburg in part fulfillment of the requirements for the degree of Masters of Medicine in the branch of Haematology Johannesburg, / Parvovirus B19, a single stranded deoxyribose nucleic acid (DNA) virus, is known to cause anaemia in the setting of immune suppression such as Human Immunodeficiency Virus (HIV) infection. It is typically associated with a severe, isolated, normochromic normocytic anaemia and reticulocytopenia. The bone marrow classically shows a pure red cell aplasia (PRCA) with absence of maturing erythropoiesis, giant pronormoblasts and a variable presence of erythroid viral inclusions. Parvovirus B19 infection is a treatable cause of anaemia using red cell transfusions, intravenous immunoglobulin (Ig) therapy and in the setting of HIV, antiretroviral therapy. In the setting of HIV infection, testing for Parvovirus B19 infection using molecular techniques such as polymerase chain reaction (PCR) are preferred over serological methods, as antibodies are either not made or are dysfunctional. In South Africa, the prevalence of Parvovirus B19 infection in the HIV infected population with severe anaemia is not known. The aim of this study was to assess the prevalence of Parvovirus B19 in a cohort of HIV infected patients with severe anaemia. The Inclusion criteria for specimens into the study included all specimens submitted for a bone marrow examination submitted for routine diagnostic workup between January 2012 and November 2013 at two academic hospitals in Johannesburg. The study population included HIV infected patients with severe anaemia, defined as haemoglobin levels <8 g/dl for men and non-pregnant women. Real-time PCR using the PrimerDesign™ genesig® Kit for Human Parvovirus B19 (Southampton, United Kingdom) was performed on DNA extracted from bone marrow aspirate slides of these patients. The Parvovirus B19 results (qualitative and semi-quantitative values) were assessed in conjunction with various Parvovirus B19-related clinical and laboratory parameters obtained from the laboratory information system (LIS). The prevalence of Parvovirus B19 in this cohort of patients was 13.3% (19/143). PCR testing was possible even in samples that were suboptimal for morphological assessment, with 36.8% (7/19) of the Parvovirus B19 infection being observed in these samples. Of note, 31.6% (6/19) of the positive samples were not requested for Parvovirus B19 testing by the clinician or pathologist, indicating that it is being under diagnosed in this population. PRCA was not observed in all Parvovirus B19 positive samples, with a sensitivity and specificity of 60.0% and 85.1% respectively. Alternate causes of anaemia were present in 42.1% (8/19) of the Parvovirus B19 positive samples, including 21.1% (4/19) of cases which showed Mycobacterium Tuberculosis infection, 5.3% (1/19) with iron deficiency and 15.8% (3/19) of cases with marrow infiltration by malignancy. This highlights the importance of excluding Parvovirus B19 infection even in the setting of alternate causes of anaemia. In patients with severe anaemia and both HIV infection and Parvovirus B19-positivity, there was no statistically significant correlation between Parvovirus B19 viral load and HIV viral load, haemoglobin (Hb) level or CD4 count. Parvovirus B19 positivity was higher than expected in HIV virally suppressed patients, with a prevalence of 18.5% (5/27). However the CD4 counts in these samples were low (<350 cells/μl), suggesting that although viral suppression had been achieved, there was inadequate immune reconstitution to mount an effective humoral response to control the Parvovirus B19 infection. Serology for IgM as a method for diagnosing Parvovirus B19 infection showed poor sensitivity (60%) but good specificity (100%) suggesting that this is an inadequate screening test in the setting of HIV infection. The Parvovirus B19 positive samples had statistically significant lower reticulocyte production index (RPI) than the Parvovirus B19 negative samples. The negative predictive value of an RPI was 100%. Although this is a retrospective pilot study, notable findings were observed. In the setting of HIV infection and severe anaemia, Parvovirus B19 infection may be diagnosed by PCR even in the following scenarios: a negative IgM serology result, no morphological evidence of a PRCA, presence of other causes to explain the anaemia and confirmed HIV viral suppression. Parvovirus B19 is a treatable cause of anaemia and therefore an important entity to exclude. The cost of molecular diagnosis of parvovirus B19 is relatively higher than using serological methods, therefore should only be performed in the correct clinical setting. In HIV infected patients with grade four anaemia (Hb <6g/dl) and a reduced RPI, these findings support the use of molecular diagnosis for Parvovirus B19 infection regardless of other clinical and laboratory findings. / MT2017