Here’s a recent paper that illustrates the sorts of weird overlaps that occur between what might look (on the surface) like different fields. Take infectious disease and oncology. Classic chemotherapy drugs are generally interfere with all sorts of dividing-cell processes, and thus hit the rapidly dividing cells harder. That’s not going to be much use in a viral infection – the mechanisms of viral replication are (for the most part) quite different than the ones involved in mitosis in a human cell, and you wouldn’t expect anything that’s effective in one field to be of interest in the other. But consider the immuno-oncology area, where you’re trying to modify the immune response so that tumor cells get attacked more productively. There you can see more of a connection, since it’s these same immune mechanisms that are involved in clearing foreign pathogens like viruses and bacteria. You are, conceptually, trying to get the immune system to treat tumor cells less like “self” and more as if they were invaders from another source entirely.
Now, as I’m fond of pointing out, the viral diseases that we have had the widest variety of therapeutic options for are HIV and hepatitis C, although the current coronavirus has been rapidly climbing up those charts. With Hep C and coronavirus we can assist a patient’s own immune system, either directly by vaccination or monoclonal antibodies, or indirectly, by interfering with the virus enough (Paxlovid, the viral protease inhibitor) to allow the immune system to get the upper hand. And these will lead to the virus getting actually cleared out, from everything we can see. But HIV is of course a tougher situation. The current therapies fight it to a standstill, but the virus itself is still present, generally in a latent form. It’s nonreplicating in that state, and very difficult to detect, but unfortunately that nonreplicating state is reversible. Latent viral infections can (and do) come back, at various times and for various reasons – shingles as a years-late consequence of earlier chicken pox infection is another example of that, but HIV unfortunately doesn’t wait that long.
Viral latency in general is not understood as well as it needs to be, not by a long shot. HIV is known to hide out in CD4+ T cells, as well as in several other tissues and cell types, but it’s the T cell reservoir that appears to be the largest – especially in patients who have been on long-term antiretroviral therapy. It is these same T cells that are the focus of a lot of immuno-oncology work, with various schemes to get them to recognize tumor cells and act on them. The big player in that space for now is PD-1, a protein that is found on the surfaces of T cells (and B cells as well) that is part of the insanely complex immune regulation of their function. In this case, PD-1 is generally a brake on T-cell function, promoting immune tolerance of one’s own tissues by turning down T cell activity. And it turns out that many tumor types overexpress a binding partner of PD-1 (called PD-L1) on their own surfaces, which gives them a survival advantage as they loudly and constantly signal to passing T cells that they’re fine, perfectly reasonable citizens, nothing to bother about, etc.
What if one were to block these proteins’ active regions on the surface of these cells? Well, there you have a whole list of current therapies that are becoming more famous all the time, monoclonal antibodies that do just that. Blocking PD-1 has thus far had greater success, and the most well-known of these drugs is Keytruda (pembrolizumab), whose development has quite a history. As mentioned last month here, the clinical trial landscape is absolutely littered with studies trying to see how these various “checkpoint” therapies can be applied with other therapies and how their range of action can be extended by making other types of tumors immunologically “hot” once you have activated the T cells by blocking PD-1.
Now we get to the HIV angle on PD-1. There are number of patients who are currently on HIV therapies who also are being treated for cancer, and a number of lines of evidence (summarized in the current paper under discussion, first link above) suggest that disturbing the PD-1 checkpoint function seems to also disturb the latent HIV virus. That would be a really bad idea if we didn’t have drugs targeted at active HIV, but we do: and if you can shake the virus out of its well-hidden latent state, you might well be able to decrease the HIV reservoir in general after hitting the now-replicating virus with the antiretroviral cocktails. This new paper presents results from a trial called CITN-12, looking at 32 patients with well-controlled HIV who underwent pemprolizumab therapy for a range of different cancer types. These were grouped into three cohorts based on their starting T cell levels, to see what difference that might make as well.
And it seems that the antibody therapy did indeed cause some latency reversal, as measured by the amount of unspliced HIV RNA that started appearing after treatment with pembrolizumab. After six cycles of treatment, the reservoir of inducible HIV in the T cells of the patients also generally increased, but in the patients that were not able to complete those cycles because of disease progression, the inducible-T-cell population actually decreased. That suggests some common mechanism linking the cancer response to the HIV one, perhaps some non-PD-1 mode of T cell suppression that made both the cancer therapy and the HIV latency reversal ineffective. There was one patient that showed a sustained increase in HIV RNA along with a complete response to pembrolizumab on the oncology side. The team sequenced his virus to make sure that it wasn’t going further and clonally expanding, but fortunately there was no evidence of a selected variant form of the virus emerging under these conditions.
Now what complicates this picture is that there was already a similar study that reported last year on dual cancer/HIV patients with another anti-PD-1 antibody, Opdivo (nivolumab). This one failed to show any effects on latent virus. The authors of the current paper advance several possible reasons for this, but also say that they cannot rule out some fundamental difference between the two antibodies, based on what regions of the PD-1 protein they bind to. This is going to have to be addressed, of course, and even these positive results with pembrolizumab are going to have to be shored up with more rigorous viral assays than the current study was able to run. So far, though, it looks like anti-PD-1 therapy might be able to reduce latent HIV in patients who have been taking antiretroviral drugs. How far this can be taken and whether the virus can actually be cleared by such means is an open question – and how to do that (dosing levels, timing, monitoring to make sure that things are going right) is a whole other set of questions as well. But the virus appears to have finally been driving from its biggest hiding place. . .
