Skip to main content


Cancer Sequencing Hype And Reality

This piece in Science says something that needs to be said louder and more publicly. If you live in the US, you’ve surely seen various cancer treatment centers talking about their personalized therapy plans, and especially how they’ll tailor things to your DNA sequence and so on. You would get the impression that we have an arsenal of specifically targeted cancer therapies, waiting for patients to get their tumors sequenced so they can be paired with the optimal treatment.

That’s not true. I wish it were, but it just isn’t. And I fear that a lot of patients and their families actually believe that it is, which can make for some bad decisions (both medically and financially). I can hardly blame them much, though, because for years now there’s been a ceaseless flow of stories in the press about targeted therapies, genomic sequencing of patients, breakthroughs, advances, new DNA this and advanced sequence that and on and on. I think it’s safe to say that the average patient does not have the time nor the expertise to keep up with the actual biomedical literature, and thus relies on headlines and advertisements for an impression of the field, and that impression is pretty far off of reality.

The Science article, which covers a debate at the recent AACR meeting between David Hyman and Vinay Prasad, estimates that only about 15% of patients total are currently even eligible (under FDA guidelines) to have their tumors sequenced in hope of matching with a targeted therapy. About one third-of those may actually benefit from the process in the end. This is not exactly what you’d expect if all you knew about this stuff was what you heard on TV. The thing is, that’s actually a great advance, because the number used to be zero, in both categories. We really are making progress, and the people who can benefit really can benefit. It’s just that there aren’t nearly as many of them as we’d like, not yet.

But I can’t pin this just on clueless news readers – even back in 2003, the head of the NCI (later head of the FDA, Andy Eschenbach) was talking about “eliminating death and suffering” due to cancer by. . .2015. We have missed that deadline. In 2006 (as that link shows), he was going on in a similar way about how we’d moved to a whole new level, everything had changed, etc. In general, whenever someone starts talking about “the next level” or something like that, I check to make sure my wallet is still in my pocket. This isn’t a game we’re working through – it doesn’t have “levels”, and in fact, not many large complicated real-world things do. Quantized jumps happen on the microscale. Up here, we’re grinding it out, and that’s how cancer therapy is advancing.

Another reason for the confusion is that people have different ideas about what “precision targeting” in oncology really is. The term got applied early on to the wave of kinase inhibitors and so on that targeted particular cellular mechanisms (rather than just being wrecking balls like some of the earlier chemotherapy drugs), and that muddied things a bit. Some of the earlier successes in this area came about because some (relatively rare) tumor types really do seem to be pretty genetically homogeneous: they occur in pretty much the same way for the same reasons. So if you target that mechanism, you’re likely to see an effect; it’s like the situation in many human genetic disorders, and it doesn’t happen that way in cancer too often. But the high-profile successes against things like GIST confused some parts the public, I think, into thinking that the different bodily locations of cancer were all like that – genetically distinct from cancer in other organs, but very similar to each other.

Would that it were so. The second level of confusion comes to the DNA-sequencing aspect, which is at least closer to reality in terms of how cancer works, if not its treatment. “OK’, this story goes, “this patient has (something)oma, so let’s take some cells and see what variety of it they have so we can give them the drug that works against that subtype”. That’s what too many people think we have now, and what in a few cases we actually do (as above). But most of the time we don’t really understand what’s going on at a cellular level, as gets proven every time another promising oncology mechanism wipes out in the clinic.

And even with those successes, there’s room to wonder how far this idea can be pushed. Here’s a piece on that from 2016 (more Vinay Prasad), wondering just how precise it’s possible to get. One big problem is that many solid tumors are a terrible mixture of cell types. They generally got that way by being genomically unstable, and that unstability just goes on and on. That’s where all this fine targeting really runs into trouble: what if a given patient’s tumor is a mixture of eight or nine different cellular forms of cancer? Well, eight or nine that you know about so far?

That’s one reason that immuno-oncology is such a hot field, because (to a first approximation) it’s saying “OK, we don’t know how all you cells became cancerous and we don’t care. We can set it up so that (for example) you all get recognized by killer T cells, and they will take you out no matter what your pedigree”. Problem is, we’re not quite to this stage for all but a very few forms of cancer. By far the most successful are blood-based cancers like leukemia, where the cells really are quite similar to each other by oncology standards, giving you a chance to wipe them out en masse. Getting that to happen for solid tumors is. . .not so easy. The more successful immuno-oncology for solid tumors, so far, is using antibodies to go after potential common mechanisms like PD-L1 that are fending off such T-cell attack, but (a) we don’t have such markers for most solid tumors, and (b) in many cases, the most we can do now, usually, is to wipe out most of of a patient’s tumors (best case) leaving behind an untouched residue of mutated cells that eventually comes back in untreatable fashion.

I’m not putting that down at all – this can buy people substantial more life, although not always. In every case, you have to see (with real-world clinical data) how much actual survival you’ve been able to achieve. That’s another big reason that we missed that absurd 2015 date, because it takes years to figure out just how well these therapies actually work in practice. It’s a slog, and I don’t see any way for it not to be a slog, and it’s too bad that that doesn’t make for a very compelling slogan.

12 comments on “Cancer Sequencing Hype And Reality”

  1. genomistry says:

    I’m glad Derek’s post highlights critical assessments of “precision” or “personal” medicine. A lot of people seem to go about parroting the phrase today like zombies.

    Precision medicine has been highlighted again for a landmark NIH study that is beginning:

    The project lead, Eric Dishman, has attributed his kidney cancer free status to the pancreatic cancer drug given to him after genome sequencing (after more than two decades of chemo):

    Is it true that we can be relatively certain it was the sequencing that led to the outcome in Dishman’s case?

    1. genomistry says:

      There are a lot of stories on Dishman’s case, but I can’t find any specific about the drug trial he was enrolled in and how/if the cancer’s disappearance was attributed to the experimental drug. Is it possible that the experimental drug in the trial is effective against many kidney cancers and not just his specific case?

    2. Mol Biologist says:

      IMO, this post is a little bit out of date. The reduced price of NGS drove the wave of accumulation huge amount of ambiguous data.
      “Genomics is at the centre of much of what we do in cancer research,” Staudt says. “Now we can ask questions in a more directed way.” Why genome instability occurs? What is an advantage of genome instability for cancer cells…….Can we stop it? Or even reverse? What is metabolic demands in this particular cancer cells? There are more questions than answers, aren’t it?

  2. oncodoc emeritus says:

    I think that it is fair to say that paid announcements on television are marketing tools not information sources. I bought several six-packs of a specific beer once, but the bikini team did not show up. Further, comments from newsreaders contain sensationalized tidbits rather than sober assessments; I still haven’t died of Ebola. There has been clear progress in several neoplasms using targetted therapies, and there will be more, but there will also be disappointments and resistant neoplasms. There will be people who have great responses and people who don’t respond despite having favorable profiles. Don’t buy hype, and, yes, somethings are actually too good to be true. Hard work is the best hope for medical science; cautious hope is the best stance for patients.

  3. David Edwards says:

    Even without being a specialist in the field, I’m interested to know how it is possible to subject to “precision targeting” via its genome, what is in effect in numerous instances a genomic quick-change artist with a huge ‘costume wardrobe’ and rampant fecundity.

    Even if you find some cells from a tumour that possess a recognisable and targetable vulnerability – determining this involving, of course, a lot of hard work – that will be of limited use if the tumour subsequently acquires mutant cells lacking that vulnerability.

    One thought that springs to my mind, as an alternative route of attack, and one that might even have occurred to actual biological researchers, is to use viruses with shifting genomes to deal with cancer cells instead. Two well-known (but dangerous) candidates exist for testing the likely success of this approach, though choosing either will open up a minefield of clinical safety issues before they move from the stage of “interesting ideas worth trying”. Vesicular stomatitis virus (VSV), chosen because it is not known to be pathogenic to normal human tissue, is already under investigation as a means of dealing with various cancers – here’s a PubMed reference to the virus being investigated as a means of dealing with colorectal cancer, and I’m aware of other research (some references thereto cited in the full text) investigating VSV as a means of dealing with deep-seated gliomas. Modifying n actual human pathogen for the purpose would be much more risky, unless of course some means of control could be implemented, such as the ingenious method used in this paper to control HIV replication in a potential vaccine application.

  4. George says:

    Firstly, let’s recognize the huge achievment that is shifting from site-of-origin to a more rigorous identification of cancer type. We wouldn’t say “lung infection” without expecting a specific identification of the infectious agent, and in time we’ll view the term “lung cancer” etc. as equally limited and anachronistic.

    Yes this is a field that marches in increments, but those increments are always in the same direction. 15% will become 40%. 40% will become 70%. 70% will be come 100% or close enough to it. The view of the Cassandras (highlighting present limitations rather than future potential for growth) obscures the reality that for those patients who do go into sustained remission as a result of these targeted therapies, the effect is quite literally life-changing. Every year the number of people affected by these cures increases.

  5. Insilicococonsulting says:

    My own prediction from 10 yrs back and one I back to this day, is that personalised genomics and therapies will have their day in the sun, but will almost always be applicable to a small set of patients. Eventually, we will find therapies that obviate the necessity for this tailoring.

    1. loupgarous says:

      As underdiagnosed as pheochromocytoma, paraganglioma and other neuroendocrine cancers have been, it’s possible that genomics will find its niche there.

      One practical consequence of insufficient understanding of cancer genomics was that my own cancer, after excision of a tumor between heart and spleen that was large enough to cause severe pain (it was growing on a ganglion) and emit a mix of obscure catecholamines (along with the usual suspects, epi, norepi, metanepi and normetanepi, for example) was that the metastases went undetected and undiagnosed until they’d colonized my liver and put me in the hospital after a three-wine birthday brunch.

      The then-“gold standard” for detecting neuroendocrine tumors, radioiodine-tagged MIBG, didn’t pick up the metastases in my liver (which, by then, had grown large enough to image by MRI). My insurance company didn’t want to shell out for the test which DID find my tumors, Indium-111 octreotate, which was too bad for me. There are at least two discrete cell lines for paraganglioma, one of which is detectable by MIBG imaging, one by octreotate imaging. There’s probably some branch of the “rule of tens” that describes the relative rarity of the tumors involved.

  6. Question says:

    Where does a drug like Olaparib/Lynparza fit into this? A victory for precision medicine and small molecules?

  7. Ron Galt says:

    I’ll never understand why I don’t have cancer given all the crap I’ve worked with over the years.

  8. Emmy Hayden says:

    Great article. It was a nice read. Much of what is said in the advertisements may be not completely true but at least it is a source of encouragement for patients who are now becoming more aware, if not fully aware, of the developments taking place in treating one of the most dreaded diseases in the world. The number is at least more than zero now, as mentioned in the article.

  9. Johnh says:

    I think the one thing missed is the benefit of the data. More sequenced cancers adds to the DB of knowledge opening the door for learning via meta analysis.

Comments are closed.