US researchers have found an antibody that hunts down prostate cancer cells in mice...
That's mice, folks. Let's review the guidelines for titling your cancer-related article: If the result is in vitro, you must use the words in vitro; if it is in mice, you must use the word mice.
Meanwhile, like all crappy popular articles, this one provides no citation. It doesn't even mention the names of any of the researchers, poor dears. But Google and I can remedy this -- this seems to be the paper:
This is, as much as one can tell from an abstract, a fine preliminary result. But don't buy any stock yet.
Not only is this in mice, but this result is about attacking human prostate cancer that has been transplanted into mice. Now, this is a darn sight more impressive than killing cells in a dish, but it's still not the same problem as killing the cells in the human body where they originally evolved. The mouse is a nude mouse; it's immunodeficient (otherwise it would reject the human cancer). Part of the reason that the antibody may be working so well -- attacking the cancer, leaving other cells alone -- is that the cancer cells aren't just cancer, but also from a completely different part of the mammalian family tree.
The other big problem with these studies is that it's one thing to be able to label "85%" of metastatic cancer cells in an animal model, but that's not necessarily a cure. For one thing, among the remaining 15% may be the line of cells that is resistant to your antibody. Now you need another antibody!
Worse, antibodies are hard to deliver to every cancerous cell. They're not especially tiny molecules. They have a relatively easy time being delivered to normal tissues via the bloodstream, because normal tissues are very carefully engineered such that every cell is within a short distance of a capillary. But tumors, having gotten where they were by randomly flipping switches on their genomes like a toddler trying to launch the Space Shuttle, are pretty poorly organized. There are bits of the interior of a tumor that are so far from blood vessels that the cells there are asphyxiating. Unfortunately, because of this, one of the things that cancer cells tend to learn really early is how to survive in such low-oxygen conditions, and when an antibody comes along in the bloodstream and kills the easily-accessed cancer cells, it is likely to leave behind some of these orphan cells, which sit around waiting for the day that they will come across some more oxygen and start growing again. Then, a few months or years later, your cancer comes back.
(This, by the way, is also a big problem with classical chemotherapy and every other cancer treatment.)
So, this result is a relatively early report from a work in progress. Such work takes a great deal of time to come to fruition.
US researchers have found an antibody that hunts down prostate cancer cells in mice...
That's mice, folks. Let's review the guidelines for titling your cancer-related article: If the result is in vitro, you must use the words in vitro; if it is in mice, you must use the word mice.
(Why? Because mice are not like humans. See my earlier musings on that subject here: http://news.ycombinator.com/item?id=838160 )
Meanwhile, like all crappy popular articles, this one provides no citation. It doesn't even mention the names of any of the researchers, poor dears. But Google and I can remedy this -- this seems to be the paper:
http://www.pnas.org/content/early/2009/12/15/0911397107.abst...
Salient facts:
This is, as much as one can tell from an abstract, a fine preliminary result. But don't buy any stock yet.
Not only is this in mice, but this result is about attacking human prostate cancer that has been transplanted into mice. Now, this is a darn sight more impressive than killing cells in a dish, but it's still not the same problem as killing the cells in the human body where they originally evolved. The mouse is a nude mouse; it's immunodeficient (otherwise it would reject the human cancer). Part of the reason that the antibody may be working so well -- attacking the cancer, leaving other cells alone -- is that the cancer cells aren't just cancer, but also from a completely different part of the mammalian family tree.
The other big problem with these studies is that it's one thing to be able to label "85%" of metastatic cancer cells in an animal model, but that's not necessarily a cure. For one thing, among the remaining 15% may be the line of cells that is resistant to your antibody. Now you need another antibody!
Worse, antibodies are hard to deliver to every cancerous cell. They're not especially tiny molecules. They have a relatively easy time being delivered to normal tissues via the bloodstream, because normal tissues are very carefully engineered such that every cell is within a short distance of a capillary. But tumors, having gotten where they were by randomly flipping switches on their genomes like a toddler trying to launch the Space Shuttle, are pretty poorly organized. There are bits of the interior of a tumor that are so far from blood vessels that the cells there are asphyxiating. Unfortunately, because of this, one of the things that cancer cells tend to learn really early is how to survive in such low-oxygen conditions, and when an antibody comes along in the bloodstream and kills the easily-accessed cancer cells, it is likely to leave behind some of these orphan cells, which sit around waiting for the day that they will come across some more oxygen and start growing again. Then, a few months or years later, your cancer comes back.
(This, by the way, is also a big problem with classical chemotherapy and every other cancer treatment.)
So, this result is a relatively early report from a work in progress. Such work takes a great deal of time to come to fruition.