On Nov. 7, 2019, a group of researchers from Abbott Diagnostics published an article in the respectable (but not typically thrilling) Journal of Acquired Immune Deficiency Syndromes. It reported the DNA sequence of CG-0018a-01, an obscure HIV strain collected in 2001 in the Democratic Republic of Congo. So far, so underwhelming.
But it turns out that CG-0018a-01 is genetically similar to two previously discovered strains of HIV—and a collection of three related HIV strains is enough to define a new family, or subtype, of the virus.
Welcome to the world, HIV subtype L, the first new HIV subtype to be discovered this century. And cue the mainstream news coverage from media outlets like CNN and the New York Post.
But even in that initial coverage, there were signs that publicity around the story may have created more impact on humanity than subtype L itself ever would. “There's no reason to panic or even to worry about it a little bit,” said Anthony Fauci, M.D., director of the National Institute of Allergy and Infectious Diseases (NIAID), to CNN. “Not a lot of people are infected with this. This is an outlier.”
This incident did, however, bring up some broader questions. What is the significance of HIV subtypes? And should people who are living with HIV, or concerned about HIV, worry about the possibility of getting a second infection with a new subtype?
HIV’s Family Tree: A Series of Unwanted Starbursts
Subtypes exist because HIV is small but prolific. Inside a single untreated individual, HIV produces more than a billion copies of itself per day. As it makes those copies, it does so in a way that is deliberately sloppy, thus generating vast numbers of variations.
The result is massive genetic variation: The percentage of genetic variation within a single HIV subtype is more than 10 times the percentage of genetic variation between humans and chimpanzees. Variation between two distinct HIV subtypes is two-fold greater again.
That so-called “starburst diversity” creates a kind of family tree for HIV, generating branches that then become their own families. The main, or M, group of HIV-1 is responsible for the vast majority of HIV infections worldwide. The constituents of group M are 10 sub-branches or subtypes, which vary by geography. The most notable are:
- Subtype A, which is most common in parts of East Africa, Russia, and the former Soviet Union countries.
- Subtype B, which is dominant in Europe, the Americas, and Oceania.
- Subtype C, which rules the roost in Southern Africa and India.
In the rest of Asia and in Western Africa, the most common strains are called “circulating recombinant forms” (CRFs)—a genetic mash-up of pieces of two different subtypes.
Subtype L is the newest addition to group M.
Why Do HIV Subtypes Matter?
All of this diversity in HIV strains matters to researchers for several reasons.
First, HIV diagnostics need to detect all of those strains—indeed, this is why Abbott Diagnostics was interested in subtype L in the first place. Thanks to previous work on understanding HIV diversity, “our testing is so good now that we can pick up all HIV-1 groups, and even HIV-2,” says Davey Smith, M.D., a virologist at the University of California, San Diego. That creates confidence in a negative test result, ideally reducing the number of people with persistent fears about HIV infection even after multiple negative tests.
Second, HIV epidemiologists use the telltale diversity of HIV strains to understand how the virus spreads through different populations, which can influence the strategies used for HIV prevention work.
Third, and perhaps most critically, any future vaccine will need to protect against all HIV strains if it’s going to be truly effective. The outside coating of HIV is the area most prone to variation, which makes it particularly difficult to make a vaccine that protects against all HIV strains.
That’s why subtype is important to scientists. But does the infecting subtype matter for people living with HIV?
With some minor caveats, the answer is no.
“Of all the things to worry about, that is not one,” says Smith. “In reality, it doesn’t matter with the current treatment”—and that is as true for subtype L as it is for the previously known subtypes.
On the discovery of subtype L, Smith says: “It’s like finding a lost cousin; it’s nice to know it’s out there, but it doesn’t matter in terms of how we test or how we treat [people who are diagnosed with HIV].”
HIV Subtypes Have Little Effect on HIV Treatment and Disease Progression
Long before there was a subtype L, HIV experts felt a moment of nervousness about what new HIV subtypes might mean.
When the first HIV medications were being researched in the 1980s and 1990s, clinical trials of antiretroviral treatment were conducted almost entirely with U.S.-based patients who were living with subtype B. But subtype B is responsible for only 12% of HIV infections worldwide.
There was a sigh of relief when it became clear that HIV treatment also worked for other subtypes, including subtype C, which is responsible for roughly half of all infections globally. “We got lucky,” says Smith.
Some HIV subtypes do more commonly have certain drug-resistance mutations, or can more easily become resistant to some types of antiretrovirals, but this can be addressed by conducting an HIV drug-resistance test before a person starts treatment. And, says Smith, “the current treatment is so good, it takes care of those issues.”
Treatment is, however, not the only issue to worry about with subtypes. “There is some evidence that some of these subtypes may cause faster disease progression,” says Smith. But he’s quick to note that, even if such a thing is true, the difficulty of finding conclusive evidence “tells me that it’s not a big effect,” he says.
Furthermore, even the limited evidence about subtypes and disease progression came from studies of people in Uganda and Kenya who were not yet on HIV treatment. Since then, immediate commencement of HIV treatment after diagnosis has become the global recommendation, and the point has become somewhat academic.
As a result of all these factors, focusing on subtypes “is not something that is in the prioritization of messaging” to HIV providers, says Hyman Scott, M.D., M.P.H., medical director of clinical research at Bridge HIV within the San Francisco Department of Public Health.
How Often Does HIV Superinfection Actually Happen? And How Dangerous Is It?
Science has shown us that the subtype of the initial infecting strain is not particularly important in determining whether HIV treatment will work or how fast HIV disease might progress. But what about the prospect of a second infection within a person who is already living with HIV?
In 2002, data presented at the XIV International AIDS Conference raised a scary new prospect for people living with HIV. Bruce Walker, M.D., a prominent clinician-researcher at Harvard's Massachusetts General Hospital, reported that a man living with HIV had been superinfected (sometimes also called “reinfected”) with a second HIV strain. A previous report of superinfection in 2000 had turned out to be a false alarm, but the Walker case was confirmed—and two other case reports were also published that same year.
Even after handling the shock of an initial diagnosis, these findings made it seem that HIV-positive people were not free from concerns that they might be superinfected with another HIV strain—and potentially one that was more virulent than their existing infection.
Andrew Redd, Ph.D., at NIAID was intrigued. “Yes, HIV superinfection does occur, but [reporting on superinfection] was all anecdotal,” he says. “I was interested in how often superinfection occurs.”
Redd looked for answers within a unique population in Rakai, Uganda. Starting in the 1990s, before antiretroviral treatment was available in the country, people in Rakai had been provided with HIV prevention messaging and monitored regularly to track the spread of HIV. Redd re-examined these historical blood samples and found in several studies that “the rate of superinfection was roughly equivalent to the rate of initial infection. What that indicated to us was that the natural immune response to initial HIV infection is, in many cases, not sufficient to protect you from a second infection.”
That finding also had huge implications for developing an HIV vaccine. If infection with a strain of actual HIV could not protect someone from future reinfection, what hope would a vaccine have?
But the more immediate question was what this meant for people already living with HIV. Some of the case studies raised concerns: At least one individual had controlled his initial HIV infection, but then struggled with a superinfecting virus that was resistant to two of his treatment drugs.
Meanwhile, several individuals had an initial infection with drug-resistant HIV, but that strain was then hidden after superinfection by a more vigorous, albeit drug-sensitive, strain. This meant that if drug-resistance testing only picked up the newer strain, these individuals may have been put on an incorrect treatment regimen that allowed their original, drug-resistant strain to flourish.
In the Mid-2000s, HIV Superinfection Fears Became Common
The results of these studies started to filter into the consciousness of both HIV physicians and people living with HIV. A study published in 2004 focused on HIV-positive gay men in San Francisco. Of those who had heard about the concept of reinfection, almost two-thirds “agreed or strongly agreed that they were concerned about reinfection.”
This awareness also filtered down into the sexual lives of people living with HIV. Compared to those who were not aware of reinfection, gay men in the study who said they knew about reinfection were half as likely to report unprotected anal sex with another HIV-positive individual. “Over half of participants strongly agreed that they were safer sexually because of concerns about reinfection,” the study authors reported.
The authors noted that these behavioral changes were occurring “despite the lack of widespread clinical evidence that reinfection leads to adverse clinical outcomes.” Such fears persisted for years: A study in 2011 found similar behavioral results despite continued uncertainty around the evidence, with interviewees responding to competing and conflicting advice.
The scientific and medical literature on superinfection emphasized risk and threat—especially from superinfection with drug-resistant viruses. There was little, if any, anticipation within the text of these articles that the growing efficacy and convenience of antiretroviral treatment might allay these concerns.
Read in retrospect, this omission might seem odd. However, the mid-2000s were a complicated time for people living with HIV and HIV physicians alike. Immediate treatment for all, and our awareness of the hugely beneficial consequences of that, were in the future and not yet widely discussed. Although major strides were being made in HIV treatment in wealthy countries, globally, HIV drug side effects were still significant and troubling, leading many to delay treatment initiation.
In addition, the HIV expert community was still engaged in vigorous debate about when to start treatment, as well as the extent to which treatment reduced HIV transmission. For those not on treatment, superinfection was clearly possible at significant rates and could lead to a burst of viral replication, faster progression in symptoms, and the risk of infection with a newly drug-resistant virus.
“We were finding that people who had superinfection were doing worse,” says Smith. Superinfection was therefore added to the list of things for people living with HIV to worry about.
PrEP, U=U, and Treatment as Prevention: A New Era for HIV Superinfection
When the good news finally came, it went largely unnoticed, at least in terms of the ramifications for superinfection. This may be because the initial clues that superinfection might be less important in the modern antiretroviral era came not from superinfection researchers, but from an entirely different source: research on HIV pre-exposure prophylaxis (PrEP).
In the run-up to PrEP’s U.S. Food and Drug Administration approval in 2012, research showed success rates of around 99% in preventing infection among those who adhered to the daily regimen. “If PrEP works in uninfected individuals, then antiretroviral therapy in a setting where you are already infected should be at least as good” at preventing infection, says Viktor Müller, Ph.D., of the Institute of Biology at Eötvös Loránd University in Budapest, Hungary.
Not long after PrEP made waves, another set of major HIV prevention studies altered the way we think about HIV transmission and risk: The PARTNER and PARTNER2 studies conclusively showed that people who are living with HIV and have an undetectable viral load are biologically incapable of infecting others during sex, regardless of whether a condom is used. We now generally refer to this fact as “undetectable equals untransmittable,” or U=U. It is also called “treatment as prevention.”
Taken together, all of these research developments made it vanishingly unlikely that two HIV-positive people on effective treatment could superinfect each other. U=U logic dictates that neither person could be infective to the other (or to anyone else) in the first place. And based on the PrEP data, neither could be superinfected anyway, due to the effective HIV drugs in their system preventing any new infection from occurring.
Müller’s study confirmed this hypothesis. Although the study design was, in a sense, biased to find superinfection—and particularly to find superinfection that led to drug-resistant virus able to fight back against existing HIV treatment—what the group discovered was reassuring. “We wanted to see if superinfection can transmit drug resistance,” says Müller. “What we found is that, fortunately, this is extremely rare.”
Müller and his colleagues relied on an existing database covering over 30,000 individuals on HIV treatment in five European countries. The database listed parts of the HIV genetic code that had been found in these individuals, based on a drug-resistance test—that is, a test that is done at the time a person starts treatment to determine which HIV drugs will be effective against their strain of the virus.
Every now and then, an HIV care provider will issue a second drug-resistance test. The repeated test “is normally only done when you have some problem—when virus shows up in the blood again,” says Müller. One potential cause of that viral resurgence is superinfection. “So it is a targeted survey of the problematic cases.” A second test was conducted for 4,425 of the study participants.
Of the 4,425 patients who did have a second drug-resistance test, only 107 showed signs of a second strain of HIV emerging. The other 4,318 had no sign of a successful superinfection.
Müller’s team dug further in an effort to confirm how many of those remaining 107 tests were indeed evidence of superinfection. They did in-depth analyses on a subset of 14 samples, re-testing them to be sure the original findings were correct. Of those 14, 12 turned out to be false hits. Müller theorizes that the original commercial laboratories who had done the testing either contaminated or mixed up samples.
In total, out of the more than 30,000 people on HIV treatment who had been included in the study, only one person was found to have an increase in drug resistance that could be plausibly linked to a possible superinfection.
Müller emphasizes that superinfection can still happen. There were genetic signs of some transient superinfections, he says—but the new virus did not persist. “If superinfection occurred, it didn’t cause a problem with drug resistance,” says Müller. “At the moment, it doesn’t seem to be a relevant risk for the efficacy of antiretroviral treatment.”
Communicating the Modern-Day Risk of HIV Superinfection
The interpretation of Müller’s study and other evidence is clear. “If you are on antiretroviral drugs, they do protect you from superinfection,” says Andrew Redd of NIAID. “Most of us are moving on and studying other aspects.”
Redd, for example, has recently collaborated on a study showing that organ transplantation between two people living with HIV is a viable approach. “We found that even though we can see the virus [in the organ recipient] post-transplant, that does not create a sustained superinfection—which is further evidence that being on antiretroviral treatment can protect you from superinfection,” he says.
Hyman Scott, the physician-researcher in San Francisco, does not even see superinfection as an important topic for doctor-patient discussions. “It’s not high on the priority list,” he says. “I answer questions about it if my patients ask about it. The answer is yes, it’s possible, but it’s not clear what implications it has for your health.”
As a clinician, agrees Redd, “you have a lot of things to go through, so getting into the nitty gritty of superinfection is probably not that useful.”
Still, the risk is not zero. Superinfection, Müller notes, “can cause low-probability events with high impact. If two viral strains meet in the same cell, the hybrid virus can be something quite new”—with unpredictable results. Indeed, this kind of genetic mixing is how the CRFs that are dominant in Asia and West Africa came about.
In addition, “There is some controversy that people with two viruses might not do as well, though they probably got those two viruses when they were not on therapy,” says Davey Smith, the University of California-San Diego virologist.
Finally, there is still the possibility of a person transmitting an HIV variant that is resistant to multiple HIV drugs, such that the new virus can escape effective control under a person’s existing HIV-treatment regimen. “Such variants are pretty hard to evolve,” says Müller. “But we can’t count on it to stay that way. It’s no time for complacency.”
A study recently published in the journal AIDS found that a selected group of gay men living with HIV in four U.S. cities—all of whom had a detectable viral load—had significant rates of HIV drug resistance to individual drugs (31%), to multiple classes of drugs (12%), and to drugs that are critical to modern first-line therapy (8% to 16%, depending on the drug).
These individuals were specifically selected for the study because their viral load was not under control, so they would be expected to have higher levels of resistance than a broader sample of gay men living with HIV. But the findings suggest how drug-resistant HIV, combined with a detectable viral load, could potentially lead to a troubling superinfection. Indeed, a small number of case studies do demonstrate the possibility of someone on treatment being superinfected with drug-resistant HIV.
But overall, there are many more common health threats than superinfection for those who are already living with HIV. “I end up talking to my patients more about syphilis and hepatitis C and other sexually transmitted infections, because that is much more of a concern for people living with HIV,” says Scott.
The San Francisco Department of Public Health emphasizes a sex-positive approach to HIV prevention options, including both condom usage and frequent testing for sexually transmitted infections. HIV-positive people who are already engaged in care also have more opportunities to receive frequent testing for sexually transmitted infections, but the recommendation applies to all those who are sexually active. Says Scott: “There is not specific messaging for people living with HIV to do something different around sexual health.”
A Twilight for HIV Superinfection Research
As access to ever more effective and less-toxic antiretroviral treatment has spread around the world, studies of HIV superinfection have largely vanished. “The field has now moved,” says Redd. Clinical priorities have also changed: “Our one message,” says Smith, “is therapy, therapy, therapy.”
As for Smith’s research, he still uses viral sequences to track HIV, but his focus has shifted from how HIV moves between people to how HIV moves within an individual. “Most of what we know about HIV is its behavior in the blood, because that is what we can get to,” he says. But the human body has many different reservoirs—places where HIV hides in the body. “To really cure this thing, we need to drive all the HIV out of these reservoirs,” he says.
Smith is therefore focusing on the “last gift cohort”—a group of HIV-positive people who are terminally ill from a non-HIV cause. These individuals give permission for a rapid autopsy after death, which allows his research team to see where and how HIV is persisting in different reservoirs in the body.
If the work is successful, it would be an interesting twist: This area of research, which started amid fears about a person getting HIV twice, may provide clues on how to eliminate the virus from the body entirely.