Hacker News Reader

Comment by aspensmonster | Hacker News Reader

This certainly seems like a very interesting and double-edged sword of a treatment. The engineered T-cells nuke the tumors, which is a good thing. But if I'm reading this right, the large amount of cytokine that gets released from such a rapid destruction also wreaks havoc on your kidneys and could itself be fatal. It sounds like an all-or-nothing treatment that you either sweat out or succumb to. Apparently there's a drug (tocilizumab) that essentially acts as an undo button by taming the engineered T-cells and avoiding such fatalities, but then you're back to square one. It looks like the Penn team has partnered with Novartis to study the engineered T-cells and (I'm guessing) develop a method to tailor the engineered T-cell growth rate to a magnitude the kidneys can manage.

View on HN

Cytokine storms are fairly well-understood complications of various infectious diseases like Spanish Flu, SARS, and Hantavirus. There's no perfect cure for them, but there are a number of treatments that seem to help, and research into that benefits a number of other deadly ailments other than this cancer treatment. This is not a phenomena unique to this treatment.

http://en.wikipedia.org/wiki/Cytokine_storm

One answer that immediately struck me was generally in medicine if something is on a decline, waiting for it to get worse is not a wise choice. On the other hand at research time, administering something that might simply kill someone or confuse them into not taking a known existing "good" treatment path because they're experimenting has some ethical issues (unless you're into patent medicine, fads, faith healing, homeopathy, crystal healing, that type of thing).

So in the long run, rather than sitting around waiting for the tumors to weigh 7 pounds because every other current treatment failed as per the article examples, this might be an early treatment option. Certainly "processing" 7 pounds of dead tumor is going to be hard on your biochemistry. But years or months earlier in the progress of the disease, a couple grams of dead tissue, maybe even less, eh, no big deal.

Gangrene of a leg, today, well that's gonna be a tough one. Minor infection of a toenail six months earlier, eh, no problemo.

The interesting part of the human interest puff piece is they ran several trials and of course only the successes get any commentary. Curious what happened with the failures, ranging from no effect to maybe the treatment killed them. Although the journalist thinks what happened to the most successful patient will define the future of this treatment, in the real world, what happened to those folks where it didn't work is likely to define the future. Compared to the amount of ink spilled about baseball, not much was mentioned, which sounds like good news...

Wondering why dialysis couldn't be employed to take the load off the kidneys during the worst of it. Or does dialysis not remove the cytokine?

I googled around looking at the opposite direction, are t-cells small enough to get flushed out by dialysis? Could not get a straight answer. There must be a lot of money to be made in dialysis; there are a lot of "popular science" type articles about it. My guess is we'd have to try a level of indirection, red blood cells aren't lost, so relative size of t-cells vs red blood cells would probably answer that.

The other idea I have is MDs are not overly dumb, so if its possible for natural or artificial kidneys to filter the bad stuff to the outside while keeping the good stuff inside, they would probably just absolutely flood the patient with IV fluid. The original article was very detailed about the experience; this didn't make the article, so I don't think they were doing it.

If this treatment proves 100% successful against a given kind of cancer, and the procedure is developed to the mass-production stage, I would imagine we'd start getting cancer. Then you'd only get the tiny initial tumors nuked and correspondingly not suffer cytokine storms when large amounts of cancer tissue dies at once.

Reply on news.ycombinator.com