Why do scientists always forget about luck?

You get the data when you get the data, but the information we get from this study would have been useful several years ago:

Persistence of Virus Reservoirs in ART-Treated SHIV-Infected Rhesus Macaques after Autologous Hematopoietic Stem Cell Transplant

The Berlin Patient-- Guy gets HIV, develops a blood cancer, gets radiation and chemo for the cancer, a bone marrow transplant from someone who lacks the CCR5 receptor-- one of the co-receptors HIV prefers to use to infect our cells. Years later, he is still HIV free.

How did this procedure work?

The factors thought to be involved in the favorable outcome of the Berlin patient following HSCT include (i) the myeloablative conditioning regimen; (ii) the donor's homozygosity for Δ32ccr5; and (iii) the graft versus host effect.

Did the radiation/chemo kill all the HIV+ cells?

Was having a CCR5-free donor key?

Did the donor cells kill all the residual patient immune cells (and any that were infected with HIV)?


This paper made things as simple as possible to see what happened--

1-- Took 'donor' bone marrow from monkeys.

2-- Infected the monkeys with HIV/SIV (SHIV).

3-- Gave the monkeys 'cancer treatment' (radiation).

4-- Gave them their own bone marrow back.

After the monkeys were taken off of antiretrovirals........ the virus came back.

So, we know 'cancer treatment' alone is not enough to kill 100% of the reservoir HIV infected cells.


Like I said, you get the data when you get the data, but this is information that would have been useful several years ago.

This paper very directly tests an idea we have only inferred from patients before-- radiation isnt enough. They eliminated the other variables. This is essential information. Good on these folks for systematically testing these questions on primates instead of their patients, but dammit, do it faster.

We have been answering these questions by testing alternative therapies on humans. We know that HIV patients who get radiation/chemo still have the virus bounce back, even with graft-vs-host responses. We know CCR5-negative donors are necessary-- thats why they kept those two guys on antiviral therapy via their cancer treatment with a CCR5-positive donor. That would have worked extremely well, if their physicians had not, bafflingly, suggested they go off the antiviral therapy. If someone had performed this papers experiments earlier, these men might not have ever gone off of treatment.

We also know CCR5-negative donors arent enough, and could, in fact, allow CXCR4 viruses thrive and kill the patient, shockingly quickly. With this outcome, Im not sure it is even ethical to try another CCR5-negative transplant, and Im not sure where it puts the scientists who have developed ways of using gene therapy to 'make' people CCR5 negative. People studying the fundamental concepts behind how, potentially, the Berlin Patient was cured, might have revealed problems like this in animal trials, and prevented this mans death.

More labs focusing on the fundamentals of how this virus works can only decrease the 'set-backs' we have seen seemingly with increasing frequency recently, in humans.


But I think studying how the virus works under these circumstances, and what contributed to curing *The Berlin Patient* are different questions.

What specifically happened in The Berlin Patient was almost certainly a combination of the radiation/chemo, graft-host responses, having a CCR5-negative donor...

And dumb luck.

The Berlin patient could, and frankly, should have, statistically, have ended up like the patient who died a year after the 'miracle' transplant... but he didnt. Which means either 1) he was lucky enough to be treated before his quasispecies expanded to include CXCR4 variants, or 2) his radiation/chemo/graft-vs-host luckily killed any CXCR4 infected cells by chance, or 3) there are CXCR4 infected cells sitting quietly somewhere in his body, and the only reason they havent reactivated, releasing CXCR4 viruses out to kill him, is luck.

We can test all of those hypotheses, but the Berlin Patient lasting this long wasnt due to thoughtful consideration of these possibilities and treatment designed to maximize his chances.

It was dumb luck.

More like this

That is, unfortunately, a perfectly good explanation. But - wouldn't it be cool if it was not "just" dumb luck?

CRISPR CAS has begun to revolutionize scientific research and there are exciting possibilities of using it therapeutically. Some groups have made AAV that are used to infect mice in order to delete a host gene. Do you think that genome editing might be a better way to delete CCR5 (or a piece of it) rather than through bone marrow transplant? Do you think this would solve the problem of having to deplete existing cellular subsets?

By Brandt Levitt (not verified) on 07 Oct 2014 #permalink

Hi Brandt!

I think the folks trying to delete CCR5 are dead in the water for HIV+ people, after the guy who got a delta32 bone marrow transplant died. They have no way of knowing whether they will do the same thing to their patients (get rid of CCR5-tropic viruses, CXCR4 thrive without the competition, kill the patient).

They might still have a viable therapy pre-exposure (CCR5-tropic viruses get through, CXCR4 dont... usually...), but then you have to get trials to approve a therapy that has no proven immediate benefit to the patient.

I feel really bad for those researchers, right now. They had a real cool idea going.

It sounds like research may eventually lead to a greater ability to handle HIV, but probably with great difficulty.

Dunno anything about virology myself, but it seems that this and other African forest sources virii (Marburg etc) are best left there, and should have been quarantined by whatever mild or wild means needed until they drop below 1:1 -- carriers each transmit to less than one other person.

Allowing this stuff to travel through the population is like picking a fight with a bear while thinking "I have a plasma gun kit in my backpack".
In the foreseeable future we can lessen the impact of HIV but it will neither be eradicated nor controlled.

By Spectator (not verified) on 11 Oct 2014 #permalink