PhD defence Tinne Gils

Supervisors

• Em. Prof. dr. Lut Lynen (ITM)

• Prof. dr. Tom Decroo (ITM)

• Prof. dr. Erika Vlieghe (University of Antwerp)

• Dr. Klaus Reither (Swiss Tropical and Public Health Institute, Switzerland)

Abstract

One in three people with HIV (PWH) presents to care with advanced HIV disease (AHD), and one in ten PWH in the community in sub-Saharan Africa has AHD. AHD is defined in adults as a CD4 count below 200 cells/µl or a World Health Organization (WHO) stage 3 or 4 condition. People with AHD have a high mortality risk, mainly from tuberculosis (TB) and cryptococcal meningitis. The WHO has recommended a package of care, including diagnostics, treatment and prophylaxis for AHD and opportunistic infections, since 2017. However, implementation of this intervention remains limited in health facilities in high HIV/TB burden settings and is absent in the community (i.e. out of health facility). Point-of-care CD4, TB lipoarabinomannan (TB-LAM) and cryptococcal antigen (CrAg) testing are available. Antiretroviral treatment (ART) and prophylaxis for common opportunistic infections can be immediately provided. The intervention is thus suitable for use at point-of-care in primary healthcare and in the community, especially in high HIV-burden settings.

As part of my PhD trajectory, I conducted a series of studies, overall aiming at assessing the feasibility of implementation of the AHD care package at community level. Five studies were embedded in two TB triage trials in Lesotho and South Africa, recruiting 1,300 individuals near-facility and 20,000 during door-to-door community-based TB screening. Lesotho and South Africa are both high HIV- and high TB-burden settings. During the studies, the AHD care package was provided to PWH enrolled in the TB triage trials.

The feasibility of near-facility implementation of the AHD care package was assessed with mixed quantitative and qualitative methods. The prevalence of AHD was high in presumptive TB-cases presenting at the facility; 34.4% in South Africa and 56.0% in Lesotho. AHD was not associated with mortality, potentially linked to the low specificity of VISITECT. Pragmatic implementation of the AHD care package was perceived as feasible, despite challenges.

A mixed methods evaluation assessed AHD care package implementation during a community-based health campaign. We found 2.9% of AHD among recruited PWH; 4.6% in South Africa and 1.2% in Lesotho. AHD prevalence was higher in males, and in PWH with a known HIV-status not on antiretroviral treatment (ART) (compared to those on ART). The implementation at community-level of the AHD care package was acceptable and could be feasible if certain conditions are met. Implementation is conditional to availability of accurate point-of-care CD4 test, prophylactic treatment, solid referral systems and dedicated staff. 

In PWH and other stakeholders in South Africa, I qualitatively explored the meaning of AHD and perceptions on what influences its development. The term AHD was often unknow to PWH, and they were unaware for need for CD4 testing. PWH often developed AHD following i) missed opportunities to (re)-engage in care, ii) emotional stress and impaired mental health, iii) alternative beliefs about medicines and health, and iv) stigma, denial and non-disclosure.

Point-of-care CD4 testing would facilitate AHD package implementation. I therefore evaluated a novel semi-quantitative CD4 test, AccuBio VISITECT CD4 Advanced Disease (VISITECT). Using data from seven countries I compared the sensitivity and specificity of VISITECT compared to the golden standard, flow cytometry. Among 1604 included PWH, VISITECT sensitivity was good at 92.7% (95%CI 90.1−94.7%), but specificity was low; 61.4% (95% CI 58.4−64.3%), and misclassification happened frequently at high reference CD4 counts. My systematic review and meta-analysis of VISITECT’s sensitivity and specificity in currently published literature resulted in similar findings. I showed that sensitivity was good overall, but specificity was suboptimal compared to flow cytometry (65.1% (95%CI: 46.5%-81.5%)) in four studies, and higher in venous blood compared to capillary blood as sample. I prospectively studied the diagnostic accuracy and acceptability of VISITECT compared to the point-of-care Abbott PIMA analyzer. Diagnostic accuracy of VISITECT was good compared to PIMA, but acceptability of VISITECT was low in healthcare workers, due to perceived low accuracy, and not producing a numeric CD4 value.

Point-of-care diagnosis of TB would be another enabler of AHD package implementation. I therefore assessed the diagnostic and therapeutic yield of urine Abbott TB-LAM and associations of TB-LAM results with TB treatment initiation in PWH with presumptive TB. TB-LAM improved the diagnostic yield 1.6-fold compared to a composite TB reference, but the therapeutic yield only 1.2-fold. Positive TB-LAM results, when Xpert MTB/RIF was positive, were negatively associated with TB treatment initiation.

Based on collected evidence, I conclude that community-based implementation of the AHD care package is feasible when main challenges can be overcome and community-led AHD management may be necessary to face current global funding challenges for HIV.