3 ms·

Having just spent 3 days at an annual NCI meeting this article sort of made sense to me. What's fascinating is that the state of the art is moving so fast in th

Having just spent 3 days at an annual NCI meeting this article sort of made sense to me. What's fascinating is that the state of the art is moving so fast in this field that many are predicting within a few years we'll be able to deliver individualized chemo-therapies. General internists and even oncologists can't keep up with the rate of progress. As an aside, it's also a space where fast moving small startups could make a big difference

maxharris 16y ago
To really get the most you can out of the article, you might want to check out Lehninger Principles of Biochemistry, by Nelson and Cox. It's a wonderful book! Glycolysis is covered in chapter 14.
carbocation 16y ago
The key is to individualize to the level of the specific cancer, not necessarily to the patient with the cancer (Osler would punch me for saying this). I say this because, for example, leukemia is not one cancer; it is a constellation of cancers with similar features. However, if you take one specific cancer from this group (chronic myelogenous leukemia of the Bcr/abl type), you can treat it with imatinib. The survival at 5 years with imatinib is 89%--an incredible success. I'm saying this to point out that we already have individualized treatments for (a vanishing minority of) cancers. As we grow to understand the molecular and genetic basis for each individual cancer, we will, one by one, be able to target these and turn them into chronic diseases like hypertension or hypercholesterolemia. Perhaps, one day, we will even cure them. The current article being discussed actually hints at a more global phenomenon, and may produce tools that we can use as add-ons to the individualized cancer treatments that we will be giving patients. I'll just repeat the two things that I always like to say in this type of discussion: 1: We need to target the molecular basis for disease. 2: We need to remember the HAART model when treating cancer, which involves targeting multiple enzymes necessary for tumor growth and proliferation (or at least target different parts of the same enzymes) in order to fight drug resistance.
- bitdiddle 16y ago
  Well as a programmer I won't dispute your assertion, but what I heard at a keynote from a very prestigious scientist is that we will actually start targeting the patient, not just the specific cancer. He cited some examples where certain known therapies that starve tumors are disastrous in a very small percentage of patients due to certain genetic differences. It was like hearing science fiction. This is certainly a great time to be in this field.
  - carbocation 16y ago
    The scientist and I don't disagree; I think we just have a semantic difference. If the cancer is genetically different from the expectation, then it requires a different treatment. So if you have a disease that looks just like CML but doesn't have a Bcr/Abl fusion tyrosine kinase, imatinib is probably not what we'd want to use. I would call that molecule-centric, and s/he might call it patient-centric. Either way, the wording doesn't matter so long as you're treating the patient with the most appropriate therapy.
    - bitdiddle 16y ago
      I see, well I'll certainly take your word for it. I really need to read more of this literature as I've been working with these folks quite a bit, and hardly know the difference between a gene and a protein.
      - carbocation 16y ago
        Are you doing bioinformatics? If so, it's a great excuse to learn more about the field, and learning more about the field will make you care more about the work you're doing--a feed-forward loop. At least, when I started doing medical genetic research, that's how things worked for me.
        bitdiddle 16y ago
        sort of, I work in description logics, they are used to build medical terminologies. NCI has used them over the years in cancer genomics. A good amount has rubbed off just hanging out with these folks but recently I've become keen to learn more.

maxharris 16y ago

To really get the most you can out of the article, you might want to check out Lehninger Principles of Biochemistry, by Nelson and Cox. It's a wonderful book! Glycolysis is covered in chapter 14.

carbocation 16y ago

The key is to individualize to the level of the specific cancer, not necessarily to the patient with the cancer (Osler would punch me for saying this). I say this because, for example, leukemia is not one cancer; it is a constellation of cancers with similar features. However, if you take one specific cancer from this group (chronic myelogenous leukemia of the Bcr/abl type), you can treat it with imatinib. The survival at 5 years with imatinib is 89%--an incredible success. I'm saying this to point out that we already have individualized treatments for (a vanishing minority of) cancers. As we grow to understand the molecular and genetic basis for each individual cancer, we will, one by one, be able to target these and turn them into chronic diseases like hypertension or hypercholesterolemia. Perhaps, one day, we will even cure them. The current article being discussed actually hints at a more global phenomenon, and may produce tools that we can use as add-ons to the individualized cancer treatments that we will be giving patients. I'll just repeat the two things that I always like to say in this type of discussion: 1: We need to target the molecular basis for disease. 2: We need to remember the HAART model when treating cancer, which involves targeting multiple enzymes necessary for tumor growth and proliferation (or at least target different parts of the same enzymes) in order to fight drug resistance.

bitdiddle 16y ago
Well as a programmer I won't dispute your assertion, but what I heard at a keynote from a very prestigious scientist is that we will actually start targeting the patient, not just the specific cancer. He cited some examples where certain known therapies that starve tumors are disastrous in a very small percentage of patients due to certain genetic differences. It was like hearing science fiction. This is certainly a great time to be in this field.
- carbocation 16y ago
  The scientist and I don't disagree; I think we just have a semantic difference. If the cancer is genetically different from the expectation, then it requires a different treatment. So if you have a disease that looks just like CML but doesn't have a Bcr/Abl fusion tyrosine kinase, imatinib is probably not what we'd want to use. I would call that molecule-centric, and s/he might call it patient-centric. Either way, the wording doesn't matter so long as you're treating the patient with the most appropriate therapy.
  - bitdiddle 16y ago
    I see, well I'll certainly take your word for it. I really need to read more of this literature as I've been working with these folks quite a bit, and hardly know the difference between a gene and a protein.
    - carbocation 16y ago
      Are you doing bioinformatics? If so, it's a great excuse to learn more about the field, and learning more about the field will make you care more about the work you're doing--a feed-forward loop. At least, when I started doing medical genetic research, that's how things worked for me.
      - bitdiddle 16y ago
        sort of, I work in description logics, they are used to build medical terminologies. NCI has used them over the years in cancer genomics. A good amount has rubbed off just hanging out with these folks but recently I've become keen to learn more.