Since I was just banging on the table (or the lab bench) the other day about how many diseases aren’t single-factor, and about how many diseases (like cancer) aren’t even single diseases, I thought this would be a good time to haul out some evidence for that. The data are here thanks to some recent papers by groups who are sequencing various tumor lines, looking for common mutations as new drug targets. (The Cancer Genome Atlas, an NIH project, is behind a lot of work in this area).
But what’s become clear, if it wasn’t already, is that various cancer lines have a startlingly wide array of mutations. Recent work from Bert Vogelstein’s group at Johns Hopkins (with a host of collaborators) and from the CGA itself now show that there are an average of 63 mutations in pancreatic cancer cells, and 47 in glioblastomas, two of the nastiest tumors around. The first impulse might be to think “Great! Plenty of drug targets to go around!”
But hold on. For one thing, even though these mutations are surely not all equal, the fact that there are so many makes you wonder about whether attacking any one of them alone can make much of a difference. And different patients can have varying suites of those mutations, so it’s difficult to imagine that going after just one or two of those targets will be enough to treat a majority of cases. This work follows up on earlier studies in other tumor lines, all of which seem to point in the same direction: patients who are currently classed as having the same type of cancer really don’t.
This won’t come as a surprise to most oncologists, who have seen for themselves the widely varying responses to current therapies. The challenge is to figure out what these various changes mean, and how to classify patients to give them the best therapy. It’s not going to be easy. Just doing the math on the possible interactions of several dozen mutations with a list of possible treatment regimes is enough to make you pause. The hope is that most patients will fall into broad categories, which will line up, more or less, with broad categories of treatment. But it’s not going to be a good fit, most likely, and even getting those approximations to work is taking a lot of time and effort. (Just think back about how long you’ve been hearing about the wonderful new age of personalized medicine. . .)
We're not going to be able to do this, either, without a second (and much harder) stage of research: figuring out why these various mutations are important. Some of them seem to make reasonable sense, but it's not at all clear what a lot of them are doing, especially in concert with each other. There's an awful lot of ditch-digging work out there waiting to be done. For now, the quotes from Vogelstein in a Nature News summary can’t be improved on, though. This is the current state of the art, and it’s up to us to improve on it:
"It is apparent from studies like ours that it is going to be even more difficult than expected to derive real cures. . . It is extremely unlikely that drugs that target a single gene, such as Gleevec, will be active against a major fraction of solid tumours”
College chemistry, 1983
The 2002 Model
After 10 years of blogging. . .
1. Ty on September 8, 2008 8:47 AM writes...
Maybe science is really holding drug discovery back in some sense. Targeted therapy won't fare better than cytotoxic agents for broader patient population. Targeted drug discovery of the past decades just won't cut it. What we need is a better, more disease-relevant phenotypic assay and, more importantly, chemists' willingness to work with fuzziness.
Permalink to Comment2. anony3 on September 8, 2008 8:59 AM writes...
Why, exactly, are we trying to cure cancer?
The question is meant seriously. We've been trying for decades and we're still a *long* way from a cure (or cures) as Derek points out. And so much of what is in the pipeline will just extend life a bit at the margin. It's like WW I trench warfare, where we have to fight for ever yard.
Meanwhile, most cancers are easily curable, if caught early enough.
Diagnostic and detection technologies from microfluidics to gene arrays to systems biology are still in their infancy. Many of these may hold the promise of catching malignant tumors far earlier than we do now, as well as telling us what tumors do need to be worried about and which do not.
Wouldn't it make more sense to divert more resources to developing these immature diagnostic technologies, in the hopes of catching far more cancers at a time when they can be cured easily? That could be a *much* easier problem than seeking a "cancer cure."
Permalink to Comment3. petros on September 8, 2008 9:05 AM writes...
The second comment reminds me of a statement I heard at a meeting a couple of years back.
Primary cancers are generally treatable with a good prognosis for most patients. But secondary cancers continue to have very poor survival rates although modern treatments can prolong life alittle
Permalink to Comment4. Wavefunction on September 8, 2008 9:10 AM writes...
I agree with the second comment; we need to devote Manhattan Project-style resources to diagnostics since many cancers are already more or less curable if caught at an early stage. It's really the late-stages that define the problems with cancer therapy.
Permalink to CommentAs far as drugs are concerned, I think we can all agree that we need to develop safe rather than potent drugs. There must already be a hundred drugs that will kill every type of cancer and also kill the patient.
5. emjeff on September 8, 2008 9:35 AM writes...
anony3;
I think that is what is happening. Better diagnostics are in large part responsible for the huge increase in 5 year survival for many cancers.
I do think that there is work to be done with this informations. It is quite possible that not all of these mutations are created equal.
Permalink to Comment6. Cellbio on September 8, 2008 9:41 AM writes...
Does it bother any one else that Vogelstein refers to Gleevec targeting a gene rather than an enzyme? Also, it targets a single enzyme? Really? I thought the belief was the efficacy is due in part to inhiiting multiple kinases.
Permalink to Comment7. Dlib on September 8, 2008 12:40 PM writes...
Biggest risk factor for cancer....Age.
Permalink to Comment8. Andrea on September 8, 2008 2:03 PM writes...
No doctor is following me after being treated for rectal cancer in 2005-2006. I was 46 at the time.
Permalink to CommentI was given oxaliplatin, xeloda, and radiation.
I did not have surgery, and so far so good, thank goodness. I don't have health insurance any longer because I can not afford it at this time.
But, I have not been contacted by any doctor, and I have had many, as to how I am doing, and that could only be helpful to other patients who will be diagnosed.
Thought this information might be helpful.
9. Eric Jablow on September 8, 2008 7:28 PM writes...
How many mutations does the average working cell have?
Permalink to Comment10. JW18 on September 8, 2008 10:50 PM writes...
We'll give the great Vogelgram a break with his semantic misstake. Also said something about treating pathways instead of specific targets, yet to treat a pathway you have to treat a specific target, but we know exactly what he means.
Permalink to Comment11. Loon E. Toon on September 9, 2008 6:46 AM writes...
While I agree that for most cancers, early detection is key, for some versions, the progression from "cancer free" to "gonna die, just not sure when" is frightfully rapid. For pancreatic cancer, this is something like 6 months (Chari, Semin Oncol 2007 34:284). The window is very tight for early detection and treatment. There will probably always be a significant number of people who will present at fairly advanced stages, and present best medical practice doesn't offer very much.
Permalink to Comment12. MikeyMedChem on September 9, 2008 12:52 PM writes...
Ty -- interesting comment, and I've been thinking the same thing of late and discussing with colleagues that we're awfully hyperfocused on target elucidation before optimization because we think we're so good at drug design, but we're really not! The question is how to optimize with "fuzzy" data...if you make a change to a compound and it displays a phenotype you're interested in, should we care what target(s) it might hit, and in what ratio? I guess as long as we keep an eye on tox and other undesirable phenotypes down the road, we might be ok. Derek's entry on antipsychotics today may be another case-in-point: We _think_ we know how some successful drugs operate, but suddenly, down the road, we discover another biological target that these drugs hit. It seems inevitable, that something new will be discovered down the road that makes things clear, or less-so. Maybe we should embrace this fuzziness? Does anyone already embrace it? If so, how do you prosecute such a program?
Permalink to Comment13. bob on September 9, 2008 5:34 PM writes...
Actually, pancreatic cancer is the worst example of your point. It's not that it's a narrow time frame between carcinogenesis and death, it's that pancreatic cancer causes almost no symptoms while it grows so it's rarely discovered until it's too large for curative surgery. It's actually one of the best examples of a cancer that better diagnostics would massively reduce the death rate for.
But I think it's easy to oversell the benefits of this approach for most other cancers. If there's one thing the huge focus on detection in breast cancer has taught us, it's that it's very easy to detect and treat with brutal medical therapy "cancers" that never would have been a true threat.
Permalink to Comment14. Tanith on September 9, 2008 8:23 PM writes...
Hi,
Just in reference to diagnostics and improved 5 year survival, you need to be careful in the interpretation. Some people may be living longer with cancer not because it was succesfully treated early but because it was detected earlier in the natural history of the disease.
Permalink to Comment15. Chemist of Sorts on September 9, 2008 8:58 PM writes...
Cancer is a great example of where biochemical understanding lacks. Regarding comment 2 though, to question whether we should even be trying to improve treatments and to say that most cancers are easily curable reflects profound ignorance and insensetivity. True enough many cancers are treatable to some extent, but few are curable. Even if a cancer is beat into permanent remission, the treatments (radiation and chemotherapy) lead to huge risks of secondary cancers after 10+ years. Just because advances in oncology are slow does not mean that we should not still be trying to make them.
Permalink to Comment16. Cellbio on September 10, 2008 12:10 AM writes...
To Ty and MikeyMedchem
Permalink to CommentHave done phenotypic screening, and it works. Screened about 30K cmpds in more than 20 cell measures, and 250K in a paired assay format. We could define fingerprints of known inhibitors and find novel chemical matter that hit the known target. Could also see biological differences in "target-based" cmpds not evident in any selectivity data. Could also define profiles of interest, like known drug profile w/ greater selectivity. In one case found a series with highly selective cell impact (1 of 22 measures). With reliable pharmacology, good curve shape and SAR, we thought we removed fuzziness, but alas, little support w/o target. Interestingly, to me anyway, chemist were earlier and fuller converts than biologist. To answer the question of how to prosecute the program, I could go on at great length, but the first lesson is to not do it in a target-based company.
17. Sane Biochemist on September 10, 2008 4:42 AM writes...
Without reading the paper, did they say how long the cell lines had been in culture. I suspect some of mutations occur in continuous culture, just look how cell some lines are adapted to be responders to specific cytokines and what machinery they dump when not needed.
On another note I once went to a lecture by one of the UKs top oncologists (some 10 years ago so I cant remember who). To quote him them " with existing slash, burn or poison therapies, we can kill all cancers, if we can get the drugs to the right cells." He was a firm supporter of efforts for drug targeting rather than new drugs etc.
Permalink to Comment18. Jack Bauer on September 10, 2008 2:01 PM writes...
I've come to the conclusion that perhaps cancer will never be cured. Don't these large companies make more money off of chemotherapy treatments then they would with a cure-all drug? If a break through does come, most likely it'll be a small research group in a town we've never heard of. And would curing cancer be great for society? I'm not trying to sound evil, but keeping every human alive as long as naturally possible seems like a double edged sword.
Permalink to Comment19. Malcolm on September 10, 2008 9:28 PM writes...
Cancer is cellular evolution at work within the ecosystem that is the body.
The enabler of evolution through natural selection is random variation, aka entropy. Good luck with fighting the laws of thermodynamics.
Permalink to Comment20. Chemist of Sorts on September 10, 2008 9:31 PM writes...
“And would curing cancer be great for society? I'm not trying to sound evil, but keeping every human alive as long as naturally possible seems like a double edged sword."
Why do people say such stupid things? No one who has had cancer or has watched a loved one with cancer would say something like this. It is not necessary as scientists to leave our humanity at the door. Cancer is an awful disease that often strikes people who are relatively young. It is also more prevalent than people realize.
I firmly believe that as med chemists we have a duty to tackle problems of medical importance and medical need. I get angry every time I see another commercial for an ED product. Maybe Pharma deserves America's cynicism. Who can blame them when them when people are bombarded by DTC ads for lifestyle drugs whose development may have come only after an antibacterial or oncology program was cut.
Permalink to Comment21. Cellbio on September 11, 2008 9:57 AM writes...
I was going to stay quiet, but now will follow Chemist with additional comments.
Curing cancer would be wildly economical compared with ineffective treatment, hospitalization, lost productivity, etc. So a therapy that "cures" cancer, or eliminates the need for cytoxics, surgery, radiation (would these really be abandoned) could command a hefty price while also lessening the total cost of care. Finally, if a company making money off of chemotherapy can envision a cure, so can those not in the chemo market, so protecting the cash cow is not a good argument for why there is no cure. It is just damn tough to beat a cell wired for growth and survival.
Permalink to Comment22. Scott S. on September 12, 2008 5:54 PM writes...
There are many cancers that just don't make themselves known early. Non-Hodgkin's lymphoma is one of them. The follicular variety is most often diagnosed when somebody notices an enlarged lymph node. When that happens, the disease is already stage III or IV.
Yet, the disease is very treatable, and many patients live many years with the disease.
I'm with the folks who say, "I don't care how it works, just that it works".
Prevention and early detection is fine, but tens of thousands of people develop cancer with no risk factors or preventative steps. Look at Dana Reeve, stricken with lung cancer, yet a non-smoker.
It will take 100 years to understand most of what there is to know about cancer. Fine. Let's not wait until we understand the last detail before we start trying to cure many more cancers.
The Gleevic example was illuminating. Whoever said it would work on multiple cancers? From day one anyone who knew anything knew it was active only in cancers with the Philadelphia chromosome. It was designed that way, claiming the title as the first 'rationally designed' cancer drug.
Permalink to Comment