How much would I love to have been on the ethics committee that was faced with this?

Researchers at the University of Pennsylvania were interested in a method of treatment for leukaemia that made use of modified versions of white blood cells.  Cells were taken from leukaemia patients and genetically modified in two ways: first, they were adapted to target the cancer cells; second, they were adapted to reproduce like crazy.  This second modification is important, because attempts to make use of the first have hitherto fizzled out as the modified cells died off.  Or, in slightly more scientific language, the New England Journal of Medicine explains that

[f]irst-generation chimeric antigen receptors had limited clinical activity, primarily because in vivo activation of the chimeric antigen receptor T cells induced only transient cell division and suboptimal cytokine production, which failed to produce prolonged T-cell expansion and sustained antitumor effects. These deficiencies were overcome by the addition of a costimulatory signaling domain in second-generation chimeric antigen receptors, which enhanced the proliferation, survival, and development of memory cells — features that appeared to be the hallmarks of successful therapy with EBV-specific T cells and tumor-infiltrating lymphocytes.

And this approach, based on preliminary results, seems to be very promising, having exceeded expectations.  The paper reporting the experiment is available here.

But how do you get the modification into the cells to begin with?  The team

used HIV-derived lentiviral vectors for cancer therapy, an approach that may have some advantages over the use of retroviral vectors.

And that, of course, is where things get interesting. We ought not to get too excited and reduce the work to an attempt to cure cancer by causing Aids (XKCD is, as ever, nice and sharp on this).  But you can see why the proposal might make people sit up more than they do in response to most other putative cancer cures.  And it’s noteworthy that the experimental protocol specified that HIV+ patients would not be enlisted:

Replication-competent lentivirus (RCL) may be generated during the CART-19 manufacturing phase or subsequently after introduction of vector transduced cells into the patient. However, an RCL resulting from the production phase is highly unlikely since elements are incorporated in the design of the vector system that minimize vector recombination and generation of RCL. Furthermore, the vector used to transduce the product undergoes sensitive assays for detection of RCL before it can be released to a subject. Nevertheless, generation of an RCL following infusion of the vector product remains a theoretical possibility. The consequences of such recombination events in subjects without a known lentiviral infection are unknown, and therefore subjects with coexistent HIV infection are excluded from participation in this study in order to minimize this possibility. The development of RCL could pose a risk to both the subject and their close contact(s), and therefore, monitoring for RCL will be conducted during the course of the trial. (p. 43)

Not being a virologist, I can’t be certain of this – but it looks like the risk from the lentivirus is pretty low.  The most that can be said is that the work represents an attempt to kill off leukaemia cells by introducing white blood cells that have been buggered about with by buggered-about HIV, so there’re quite a few links in the chain.  But the risk is also not likely to be nil – not least because genes are capable of doing funny things, and the whole point of research is that outcomes are uncertain anyway.  So there will have been some non-zero risk associated with the use of an HIV-derived lentivirus being put to work in a therapeutic setting.  And since HIV is generally taken to be pretty serious, that non-zero risk counts for quite a bit.  You can see why the ethics committee might have had quite a lot to think about.

But this research seems to have been overwhelmingly worth doing – and not just because it has apparently been vindicated by early results.  (You can’t, after all, justify a research programme by appealing to the results, since the justification has to come before the programme starts.)  Indeed, I wonder whether it might have been permissible even if the risks associated with the lentiviral vector were much higher.  So here’s a couple of thought-experiments.

Let me up the stakes massively: let’s imagine that the vector used was not an HIV-derived virus, but was HIV itself – or, at least, an attenuated form.*  There would be a high risk of HIV infection incurred by the recipient of the therapy: maybe not 100%, but very high.  Would research like this still be permissible?  Well, maybe not: but its impermissibility may not be as clearcut as all that.  There’s a couple of reasons for this.

First, a high-risk intervention like this would only be considered in the most serious cancer cases – cases where, to be frank, the patient is so ill that there’s nothing much to lose.  From a treatment point of view, that’s potentially quite liberating.

But the more important reason is the second, and it is that HIV is not the problem it once was.  Though neither a cure nor a vaccine is not presently available, there are treatments that mean that HIV is no longer necessarily a deadly disease.  Proper management can turn it into a chronic condition, not a deadly one.  Moreover, the life of HIV+ people who do manage their infection properly need not be significantly worse or significantly shorter than anyone else’s.  There’s probably no getting around the fact that it’s better not to be HIV+; but it’s also the case that there are much worse things that can happen to you.

Things like aggressive leukaemia, for example.

So here’s the gauntlet: imagine that you have the “right” kind of leukaemia, and that you’ve exhausted all the other treatment options.  Your oncologist tells you that there is an experimental treatment available – but the downside is that you’ll probably end up HIV+.  Would it be worth taking the treatment?  20 years ago, maybe not.  Today?  I can’t help thinking that it might be.

And if it might be worth choosing today, why wait until the other options have been exhausted?  Chemotherapy and radiotherapy are nasty; all else being equal, you’d want to avoid them.  Given the availability of antiretrovirals – and the good chance that they’ll be increasingly powerful and efficient over the coming years – at what point would it become rational to risk HIV infection as a means to treat the cancer?  And how big a risk would be sensible?

*I don’t know if there is such a thing as an attenuated HIV virus; but since we’re in the realm of thought experiments now, I don’t think it matters.  I don’t think that there’s actually a runaway trolley and a fat man on a bridge, either, but it doesn’t make the thought experiment less valid.