The Immature Platelet Fraction in HIV Patients with Thrombocytopenia.

Thrombocytopenia is a common clinical feature of HIV infection. Given the number of possible etiologies of thrombocytopenia in a patient with known HIV, a peripheral blood test effective in determining the likely pathophysiologic basis of the thrombocytopenia would be a valuable clinical tool. Immature platelets are released early from the bone marrow in response to increased platelet turnover. These platelets contain residual megakaryocyte mRNA and have been termed reticulated platelets. A new assay, the Immature Platelet Fraction (IPF), measures the reticulated platelet count in peripheral blood. Patients with increased destruction of platelets from such conditions as ITP consistently have a higher IPF percent, while patients with decreased platelet production have a low or normal IPF percent. The goal of our study was to determine the performance characteristics of the IPF assay in HIV patients with thrombocytopenia and to see if the IPF percent could be used to help elucidate the etiology of low platelet counts in this group of patients. All adult patients admitted to the Johns Hopkins Hospital with a diagnosis of HIV and a platelet count less than 150,000 were eligible for enrollment. 62 patients were identified from February 2007 to June 2007. 34 control samples were obtained from inpatients with HIV who were not thrombocytopenic. In addition, 81 samples were available from non-HIV historical controls with normal platelet counts. The mean platelet count in the HIV thrombocytopenic group was 92,000 while the mean platelet count in the HIV control group was 254,000 (p value <.001). The mean platelet count in the non-HIV historical control group was 274 (p=.34 when compared to the HIV control group). The mean IPF percent in the HIV thrombocytopenic group was 10.2% as compared to 6.8% in the HIV control group (p=.001). The mean IPF in historical non-HIV controls was 3.1% (p<.001 for both the HIV thrombocytopenic and the HIV control group). Univariate analyses were conducted to identify potential individual predictors of a high IPF percent. Backward selection was then performed using multivariate linear models with a threshold Wald test p-value of 0.05. ITP, diabetes mellitus and cirrhosis were significantly associated with a higher IPF percent with a co-efficient (95% confidence interval) of 6.98 (3.05–10.91), 4.73 (1.39–8.06), and 14.18 (9.7–18.66), respectively. CD4 count, HIV viral load, hepatitis C and reticulocyte count were not correlated with IPF percent. Our results suggest that patients with HIV have increased platelet turnover as compared to patients without HIV. Thrombocytopenic patients with HIV have increased platelet turnover relative to both non-thrombocytopenic HIV patients and to historical non-HIV controls. History of ITP, diabetes mellitus, and cirrhosis are predictive of an elevated IPF percent. Reticulocyte count is not correlated to IPF percent, suggesting that a low reticulocyte count is not a reliable marker for decreased bone marrow production in HIV thrombocytopenia. It is unlikely that the IPF assay alone can be used to determine the pathophysiologic basis of thrombocytopenia in any single patient with HIV. Further work needs to be done to clarify the utility of the IPF assay in this group of patients.