Using HIV-ASSIST to Integrate LTBI Treatment and ART in People With HIV

Last summer, a long-term occupant of a homeless shelter in San Francisco was diagnosed with active pulmonary tuberculosis. As part of the city’s public health response, all other occupants of the shelter were tested for latent tuberculosis infection (LTBI). Of these, 4 were my primary care patients. As the Clinical Director of HIV Homeless Outreach & Mobile Engagement (HHOME)—a street-based intensive case management and primary care program designed to serve people experiencing homelessness with HIV/AIDS, psychiatric co-morbidities, and substance use disorders—I am accustomed to working creatively with our dynamic multi-disciplinary team. Even so, addressing a high prevalence of LTBI in our complex patient population raised challenges.

How do LTBI treatment recommendations differ among persons with and without HIV?
In the general population, the lifetime risk of developing active tuberculosis is 5% to 10% following initial infection; however, among people with HIV (PWH) not receiving antiretroviral treatment (ART), the annual risk is 10% and increases with declining CD4+ cell counts. Thus, prompt diagnosis and treatment of LTBI in PWH is critical.

Within a few weeks of extended exposure to TB in the homeless shelter, all 4 of our patients with HIV agreed to contact testing with an interferon gamma release assay (IGRA). One patient had a negative result and the other 3 tested positive for LTBI.  I would like to discuss the medication treatment of one of the latter patients.  But first, I will introduce some general considerations for drug therapy of LTBI, and in particular, the issues that arise in persons coinfected with LTB and HIV.

LTBI treatment options in PWH
There are now numerous treatment options for LTBI in the general population, including various combinations of isoniazid (INH), rifampin (RIF), and rifapentine (RPT), which range in duration from 9 months to as little as 1 month. However, the guideline-recommended options for LTBI treatment among PWH are somewhat limited by the numerous drug-drug interactions between certain rifamycins and ART regimens, as well as the paucity of clinical trial data on shorter-course LTBI regimens in PWH.

Historically, the most robust data on TB preventative therapy among PWH are for the use of INH, as INH has no significant drug interactions with modern antiretroviral (ARV) medications. However, the efficacy of 9 months of daily isoniazid (9H) treatment is diminished by high non-completion rates—ranging from 47% to 53% in US TB programs. Thus, in a patient population such as ours, in which there are multiple logistical barriers to long-term medication adherence, 9H may not be an effective option.

Shorter-course rifamycin-based regimens, such as 3 months of once-weekly INH and RPT (also known as 3HP), 4 months of daily RIF (4R), 3 months of daily INH and RIF (3HR), or even 1 month of daily INH and RPT (1HP) may be more effective. Each of these regimens is considered  a “preferred” or “alternative” regimen for LTBI treatment in PWH by at least 1 set of global or US guidelines.  Notably, the US National TB Controllers Association and the CDC now recommend shorter-course rifamycin therapies over INH for most individuals. However, if considering one of these regimens in PWH, attention must be paid to the various drug-drug-interactions between rifamycins and ARVs.

HIV-ASSIST: Resource to support clinical decision-making for providers caring for PWH 
When considering LTBI treatment for PWH, it is wise to consult relevant clinical guidelines and studies. However, searching and reviewing the primary literature is not always feasible in the context of a busy clinical practice.  And in some situations, the various clinical guidelines—including those developed by the CDC, WHO, and DHHS—are even at odds with one another. In such settings, I recommend HIVASSIST.com as a useful clinical resource.  

HIV-ASSIST is a free, online, clinical decision-support tool that empowers clinicians to select rational, individualized ART for their patients.  It employs multi-criteria decision analysis to integrate evidence about ART effectiveness with virus- and patient-specific factors, such as HIV resistance mutations, tropism, viral load, CD4+ cell count, HLA-B*5701 status, comorbidities, comedications, prior and current ART regimens, pregnancy status, and adherence patterns.  Clinicians can input the known factors, and HIV-ASSIST outputs a ranked list of potential ART regimens.  Each ranked regimen comes with a weighted score (lower is better, indicating stronger evidence) and is linked to concise, easy-to-digest educational summaries that reference relevant clinical guidelines and primary literature.  As an online tool, HIV-ASSIST has the advantage of being updated frequently in response to guideline changes or new data, including studies on emerging treatments and drug pharmacokinetics.

I would like to share how HIV-ASSIST can be used to quickly access important clinical information pertinent to this case.

The first step is to enter patient-specific factors into the HIV-ASSIST tool (in this case, a CD4+ cell count of 590 cells/mm³, an undetectable viral load, the desired comedications INH and weekly-RPT, a preference for oral medications, and the current ART regimen—BIC/TAF/FTC).  The next step is to review the list of potential ART regimens output by the tool.  Recall that lower scores are more preferred regimens, with color-coding that visually highlights the level of evidence for each regimen (eg, dark green, light green, and light yellow represent a continuum of clinical evidence from strong to moderate). 

The patient’s current regimen, BIC/TAF/FTC, did not appear on HIV-ASSIST’s list of potential regimens when 3HP was entered in the comedications field. Clicking on the Additional Information tab and selecting Co-medication Interactions revealed the supporting information: coadministration of RPT and BIC should be avoided, due to the possibility of significantly decreased bictegravir concentrations, and coadministration of rifapentine-weekly and TAF is not recommended by DHHS due to the possibility of decreased TAF levels (“studies are planned”). In the latter case, HIV-ASSIST also provided more nuanced data from a recent pharmacokinetic study on coadministration of TAF/FTC and rifampicin (Cerrone, et al., CROI 2018).

The 3HP regimen is a preferred LTBI regimen and an excellent fit for this patient’s lifestyle, but not recommended with this patient’s current ART regimen. So, we elected to modify his ART regimen for the duration of his LTBI treatment.  The 4th-ranked regimen by HIV-ASSIST, DTG + TDF/FTC, is one that we considered a good option based on this patient’s overall social and medical circumstances.  As was noted in the HIV-ASSIST educational information, DTG is not currently a guideline-recommended ARV in combination with RPT, due to increased DTG clearance and a theoretical risk of treatment failure or developing resistance.  But HIV-ASSIST also provided information from the recent DOLPHIN study (Dooley, et al., 2020) that demonstrated favorable results for patients who maintain excellent adherence to DTG while receiving once-weekly RPT in the context of 3HP. In this study, although DTG concentrations were reduced, they remained above 90% of minimum inhibitory concentration in all but 1 individual. Our patient had stable viral suppression and excellent adherence, so we felt comfortable using a DTG-based regimen during his 3HP treatment. We also felt that the risks of TDF-associated bone or renal toxicities in our patient were minimal, and so elected to switch TAF to TDF to minimize the potential interactions with RPT.

In summary, our selection of DTG + TDF/FTC for the duration of LTBI treatment with 3HP was based on a comprehensive clinical assessment, supported by the easily-accessible clinical guidelines and data in HIV-ASSIST. Rather than suggesting a single “right” approach to ART selection, HIV-ASSIST provides clinicians with the relevant information they need to make their own data-driven decisions.

Your Thoughts?
How would you have approached managing this patient with HIV following an exposure to TB? Answer the polling question and join the conversation by posting in the discussion section.