The impact of HIV infection when superimposed on pulmonary tuberculosis (either active or sequelae tuberculosis) on the success of bronchial artery embolisation.

Abstract

ABSTRACT Pulmonary Tuberculosis (PTB) rates in Kwa Zulu Natal (KZN) is amongst the highest in the South Africa and is often associated with Human Immunodeficiency Virus (HIV) co-infection. Bronchial Artery Embolisation (BAE) is an expensive, time consuming procedure requiring operator skill and is accompanied by risk to both patient and operator. Aim: To investigate the impact of HIV infection when superimposed on PTB (active or sequelae) on the success of BAE. Method: A retrospective cross sectional study with descriptive and analytical components of the BAE procedure between January 2006 and December 2007 was performed on sequential BAE studies. These were analyzed for procedural and clinical outcome and reasons for procedural failure were investigated. The impact of CD4 level on procedural and clinical failure was investigated for a subset of cases. Cases were included if they presented with massive or life threatening haemoptysis with a diagnosis of previous or active PTB (made clinically, radiologically or microbiologically) in whom HIV status is known and where the clinician assessed a need for BAE, but excluded any third or more attempt at the procedure for that patient. Results: The final sample size after exclusion of 91 cases is 107. Each attempt at BAE was viewed as an individual case. The study population is made up of 74 HIV positive and 33 HIV negative cases. The median CD4 level is 176 cells / microlitre. Statistically, procedural success does not imply clinically successful outcome.HIV status does not correlate significantly with clinical or procedural results of BAE.CD4 level does not correlate significantly with clinical or procedural results of BAE. There is no technical reason of statistical significance that impacts on the success of the procedure when correlated with HIV status. These include being unable to select, unable to subselect, unable to engage securely, reflux, presence of fistulae and the presence of spinal feeder arteries. The complication rate is not statistically significant when correlated with HIV status. The differences in follow up of clinically unsuccessful cases were not significant when correlated with HIV status. On imaging, all cases demonstrated pathology. No particular zone is significant when correlated with HIV status. The most common finding is parenchymal architectural distortion followed closely by features of active tuberculous infection and no statistical significance is attributed to either when correlated with HIV status. The detection of lymphadenopathy is noted in 19.1% of HIV positive cases and 42.4% of HIV negative cases, and is the only feature of significance when correlated with HIV status. Interpretation: Coinfection with HIV does not have an impact on the success of BAE in patients with active or sequelae PTB who present with massive or life threatening haemoptysis. The rate of technical failure of the procedure suggests that this needs to be performed by persons that are adequately trained. Technical success does not imply clinical success but this finding was not statistically significant when correlated with HIV status. Re-evaluation of the procedure technique and improvements in local practice may produce results that correlate better with international standards.