HIV and ID Observations

Plaza Mayor, Madrid. 1900.

From the start of the COVID-19 pandemic, one of most common questions I’ve received has been whether COVID-19 has different clinical manifestations in people with HIV.

Would it be more lethal since people with HIV have impaired immune systems? Or milder since some of the damage in severe cases is immunologically mediated?

Or would it be similar, since antiretroviral therapy (ART) is so effective?

Or maybe people with HIV are protected, as they are already taking antiviral medications — some of which (in particular the nucleoside reverse transcriptase inhibitors, NRTIs) have activity against SARS-CoV-2 in experimental systems.

The honest answer, as it has been for so much of this new disease, is that we don’t know definitively — but the early evidence from small studies suggested COVID-19 was quite similar in those with and without HIV.

This passed the anecdotal test too. Here in Boston, all of us HIV/ID specialists have had patients with HIV develop COVID-19. Absent other medical problems known to be associated with poor outcome, they mostly did fine.

What we need, of course, are larger, well-conducted cohort studies to confirm these impressions. And those are just starting to appear.

In the Annals of Internal Medicine, Spanish researchers describe the incidence of COVID-19 among 77,590 HIV-positive persons receiving ART, looking also at their risk of hospitalization. They primarily focused on their antiretroviral therapy — in particular, the NRTIs.

During a 3-month period, 236 people with HIV were diagnosed with COVID-19, and 151 were hospitalized. The risk of hospitalization by NRTI treatment per 10,000 patients was lowest for TDF/FTC (10.5), while other NRTI strategies were similar (TAF/FTC 20.3, ABC/3TC 23.4, single or no NRTI 20.0). None of those on TDF/FTC died or were admitted to the ICU. The group receiving TDF/FTC also had a lower overall incidence of infection.

What might explain this apparent protective effect of TDF/FTC? As noted above, and further elaborated upon in the paper’s discussion section, NRTIs demonstrate in vitro antiviral activity against SARS-CoV-2; furthermore, tenofovir may have beneficial immunomodulatory effects. The higher plasma and extracellular concentrations of tenofovir DF over tenofovir AF could explain the differences between the two.

But before we leap to that conclusion, remember that similar in vitro observations exist for numerous anti-infective compounds, from A-Z and beyond. (Remember hydroxychloroquine? Oh yeah, that was a thing.) Citing these mechanistic explanations hardly proves that they do anything in human beings.

And of course the people remaining on TDF/FTC today are least likely to have many of the medical comorbidities associated with worse outcome in COVID-19 — they’re healthier at baseline. Most older people with HIV, in particular those with renal or cardiovascular disease, now receive either TAF/FTC, or increasingly, a regimen that does not include either tenofovir or abacavir. The paper does not include data on these factors.

So consider the data from this fascinating paper to be hypothesis-generating rather than conclusive — as readily acknowledged by my long-time friend and HIV/ID colleague from Spain, Dr. Jose Arribas, also a co-author on the study.

We await confirmatory observations elsewhere, perhaps in people currently taking TDF/FTC for HIV PrEP — could it amazingly also be preventing a second viral infection? — or even better, from the results of a pre-exposure prophylaxis randomized study, now ongoing.

Of course these data about the NRTIs don’t get at the original question posed at the start of this post — do people with HIV who get COVID-19 do better, worse, or the same as those without HIV?

Tucked away in the discussion section is this sentence:

In line with the greater all-cause mortality of HIV-positive persons compared with the general Spanish population, we found greater age- and sex-standardized mortality from COVID-19 in HIV-positive persons (3.7 per 10 000 compared) than in the general population (2.1 per 10 000).

A large South African study, furthermore, recently reported the following:

Large study of clinical outcomes of #COVID19 in South Africa demonstrates HIV associated with increased risk of death. Effect has not been seen in smaller European, USA cohorts thus far. No diff by viral suppression. H/T @AntonPozniak https://t.co/TXwlJzsOnQ pic.twitter.com/FUxz5J7Aog

— Paul Sax (@PaulSaxMD) June 17, 2020

The linked slide presentation contains scant data about the HIV population — only that this effect was seen in both those with and without HIV viral suppression. Apparently we’ll hear more at the 23rd International AIDS Conference in July.

So what can explain these apparently negative outcomes of COVID-19 in people with HIV, both in Spain and South Africa?

In Spain and other developed countries, I’m betting on comorbidities — hinted at by the senior author of the Spanish study.

A significant fraction of our older HIV population endured years of uncontrolled viremia, immunosuppression, and toxic ART. This group experiences excess non-HIV medical problems comparable to HIV-negative people who are 5-10 years older. Many of these issues are well-defined risk factors for severe COVID-19, and perhaps controlling for these conditions will yield a prognosis comparable to those without HIV.

For South Africa, additionally, people with HIV tend to come from much more socially disadvantaged backgrounds — it is strongly associated with poverty and lack of access to care, also  negative predictive markers in COVID-19 worldwide.

So while data from these larger studies are welcome, we eagerly await more. And in the meantime, we can contemplate what all the acronyms stand for in the alphabet soup that is Jose’s twitter profile.