Viruses and Prostate Cancer

No, Im not talking about XMRV ;)

Im talking about using genetically modified viruses to treat prostate cancer :)

Broad antigenic coverage induced by vaccination with virus-based cDNA libraries cures established tumors

First, just to be clear, I dont know how this works. :-/

I mean, I know how this 'works', but I dont know how its actually working.

Firstly, it works nothing like the anti-tumor vaccines I have written about previously. In that case, scientists used HSV-1 as a Kamikaze-- the virus delivers a protein to the cancer cell that attracted immune cells to kill the tumor.

Not how this prostate-cancer vaccine works.

They took tissue from a normal human prostate, and generated cDNA (you just want DNA versions of all the RNA the cells is making-- not every cell is making all the same RNA/proteins). They then put that cDNA into Vesicular stomatitis viruses (VSV)-- Every VSV had a different cDNA artificially inserted into its genome. So, when that virus goes on to infect a cell, it will make all the VSV proteins and it will make the little bit of normal prostate protein due to the cDNA in the viral DNA. The prostate proteins will then be presented in the infected cells MHC I molecules-- 'normal' in an inappropriate context can generate an immune response... sooo... youre basically vaccinating against 'prostate'.

The mice that got the anti-prostate vaccine lived longer/controlled tumors better than the PBS mice.

And this vaccine is not causing autoimmunity when it is injected intravenously. But it does induce autoimmunity if you inject the vaccine directly into the prostate.

I dont know how that works.

:-/

I mean, if they were doing this starting with a prostate tumor instead of a normal prostate, okay. But how the heck is a vaccine with a normal prostate clearing prostate tumors and not killing the normal prostate tissue??? And theyre doing this via a prostate-specific Th17 response (whiiiiich is usually indicative of autoimmunity) and CD4 T-cell response (think antibodies) instead of CD8 (cytotoxic T-cells) or Natural Killer cells, the cells who normally clear tumors.

I dont know how that works.

:-/

But 80% of the mice who got nine rounds of this vaccine cleared the tumor.

And then, they could make new vaccines from the tumors that were not cleared, and then some of the mice that couldnt totally clear the tumor with the generic vaccine, cleared it with a 'personalized' vaccine, and the ones who still got tumors, the vaccine 'resistant' tumors grew slower.

So, I am going to keep an eye on this, but I cant imagine a vaccination strategy like this is going to be in clinical trials, like, tomorrow. While under normal circumstances, I wouldnt need to know exactly how something was working before I believed it was working (eg the HPV vaccine kicks ass, but we still arent 100% how it is so effective), but in this case, I would need to know more about how this strategy is actually working before I would think it was safe enough for human clinical trials. I would be really, really concerned about autoimmunity. And they do not mention what they would use for human clinical trials anyway (they used human prostate cDNA in mice, which worked better than mouse prostate cDNA in mice-- what would we use in humans? mouse prostate? chimpanzee? dog?).

Its cool, keep an eye on it, but temper this idea with a bit of 'hmmm...'

More like this

I read about this on HuffPo yesterday, and I came away thinking "Huh, what? How can this be working?" and figured it was a horrible journalist given I have near zero respect for their "health" reporting. I guess it really is that weird.

Can a tumor cell overexpress a regular protein, especially on the surface? That way, if you train the immune system to attack all visible prostate proteins you might primarily hit the tumor and not the normal prostate cells.

This is odd.

Viruses injected IV almost entirely end up in the liver (the resulting liver inflammation is what killed some of the people involved in gene therapy trials), and to a lesser extent, in the spleen. Injected into the prostate would have led to infected prostate cells. Why one would vaccinate, while the other wouldn't, is a mystery to me.

One *possible* cause may be that in the liver/spleen you would get expression in the presence of reduced numbers of normal regulatory cells (which when they recognize tissue-specific antigen tend to be concentrated in the lymph nodes draining those organs). A resulting low-level autoimmunity may be enough to vaccinate against cancer.

This would be excellent if it works reliably and safely; you could vaccinate people long before they were at risk of prostate cancer!

But its odd, very, very odd.

But how the heck is a vaccine with a normal prostate clearing prostate tumors and not killing the normal prostate tissue???

Maybe it is killing normal prostate tissue. Just like surgical prostatectomy does. Except that this wouldn't do ugly things to the urethra and nervous tissue that a prostatectomy damages or removes altogether.

Speaking as someone who, if I live long enough, will develop prostate tumors (and whose sweetie's father died of metastatic prostate cancer in part because he delayed even considering the possibility): I can live without a prostate. It's not doing me any good to speak of, actually -- except that the surgery to remove it is pretty much guaranteed to cause impotence and has a high risk of incontinence.

Them, I'd rather not have. Oh, if the choice comes down to no more erections or not living to watch my grandkids grow up the choice is pretty easy -- but better to have a good third alternative.

By D. C. Sessions (not verified) on 21 Jun 2011 #permalink

I think I might have some explanations for you, but they started getting really long so I decided to write up a full post on it.

Short version:

a) I don't think the "autoimmunity" with local injection is really autoimmunity. They inject the same amount of virus (1 x 10^7 PFU) in both cases, but that means the intravenous injections are way more dilute in any particular tissue. They never actually show immune responses directed at healthy, uninfected prostate tissue.

b)

theyre doing this via a prostate-specific Th17 response (whiiiiich is usually indicative of autoimmunity) and CD4 T-cell response (think antibodies) instead of CD8 (cytotoxic T-cells) or Natural Killer cells

They say this in the text, but look at the survival curve in 3h. That sure looks CD8 dependent to me. I keep thinking I must be missing something, but I've read that section over and over and have no idea how to match the text with the figure.

c)

But how the heck is a vaccine with a normal prostate clearing prostate tumors and not killing the normal prostate tissue???

I think you answered your own question:

they used human prostate cDNA in mice, which worked better than mouse prostate cDNA in mice

I think Kevin's thinking along the same lines as me. A "feature" of essentially all tumours is that the DNA repair mechanisms get all messed up, so mutation rates go through the roof. Perhaps within all that mutation that's going on, one or more of the proteins expressed within the tumour starts looking a bit more like the human equivalent?

This isn't as far-fetched as it might seem. There's still natural selection going within a tumour - most of the more fundamental genes still have to work, after all, so a lot of the mutations will be constrained within the same general landscape as in "macro" evolution.

Call it "micro-convergence" if you like.

@ Tristan - That "micro-convergence" idea is exactly what I was thinking, but didn't articulate it well. Nice.

So, here's a dumb-ass question.
Gene drift causes quasispecies in HIV, and rapid evolution/drug resistance in cancer. Are there any anti-HIV or cancer drugs that reduce the gene drift? For HIV it seems like it would be 'easier' to find a protein/enzyme/whateves that decreases its error rate vs cancer, but maybe a drug can effect both.
If HIV/cancer are moving targets due to drift, hammering a nail in their shoe to hold them down while the rest of the immune system catches up sounds like a good idea.
Instead of looking for a short term reduction in the viral load due to a potential drug, you'd have to look for a longer term effect, or monitor the quasispecies to see if they change slower.