An "expose" full of sound and fury, signifying nothing

Due to annoying stuff at work and good stuff personally, I didn't have time to grind out my usual bit of Insolence, either Respectful or not-so-Respectful, today. Fortunately, there is a long history on this blog, full of good stuff that I can repost. So, as I did when I went to TAM7, I'm picking a couple of posts for today that originally appeared in August. This one happens to have first appeared in August 2006; so if you haven't been reading at least three years, it's new to you (and if you have, I hope you enjoy it a second time).

I'll be back tomorrow.

I debated for a while about whether or not I wanted to comment on this one. The reason, of course, is that, to some extent, I've commented on a similar article before. Also, given the utter contempt that the blogger who posted this series holds me in and his delusion that I am somehow "obsessed" with him, I worried that commenting about this series, which he posted with some fanfare, might feed that delusion some more. In the end, though, because the series was by someone other than this particular blogger and because, as before, it was presented as a stinging indictment of our system of cancer research in this nation, promising a proposal for a radical departure from what we are doing now, I decided to wait until the whole series was finished and then comment. Part of the reason is, once again, that this "expose" is being billed as some amazingly insightful indictment of the cancer establishment, when, quite frankly, it's well within the range of "conventional" thinking about cancer. In essence, it's old ideas tarted up a bit.

Yes, George L. Gabor Miklos, Ph.D. and Phillip J. Baird, M.D., Ph.D. are at it again. Like last time, they're launching broadsides at the cancer establishment while a certain blogger with a much higher opinion of his scientific knowledge than any objective measure of reality would justify gives them a fairly highly trafficked forum for their series and cheers them on from the sidelines.

Now, as then, I'm underwhelmed. In some parts, they make some halfway decent points, but it all builds to a crescendo that is anticlimactic, to put it kindly.

The series starts out with an introduction by Dean Esmay, the aforementioned blogger, who references a New York Times article on uneven cancer care in the U.S. Since I've already commented extensively on this article, I see little need to say more. So instead, let's launch into the first segment by Miklos and Baird, with the melodramatic title, Cancer Cures and Blockbuster Drugs: Who Can Handle the Truth? Despite the rather silly title (oh, my God, they're going to tell us The Truth About Cancer Research!), this is one of the less silly pieces in the series, discussing as it does the high cost of cancer care. However, even so, Miklos and Baird can't resist pouring on the melodrama:

Dina Rabinovitch, author of Take off your party dress; when life's too busy for breast cancer, reveals the day to day traumas of a breast cancer patient with an advanced form of the disease (1). The cancer has spread to other organs and her third different drug treatment is underway. It began with chemotherapy and intravenous doses of the blockbuster drug Herceptin (2) and when the cancer returned, Omnitarg was prescribed. Now the regimen is; five pills per day of the latest miracle drug Tykerb, plus eight daily tablets of the chemotherapeutic agent Xeloda every two weeks out of three, plus two morphine tablets and a diclofenac every morning and evening.

She states bluntly; "My cancer keeps recurring. Nobody can tell me why. I did the genetic screening and I don't, apparently, carry the faulty genes. So angry and increasingly so cynical about these doctors in whom I have to put complete trust."

Miklos and Baird then ask:

When it's all said and done, where are all the promised cancer cures since Richard Nixon signed the National Cancer Act into law on the 23rd of December 1971?

Doctors who take care of real cancer patients (unlike, apparently Miklos and Baird, neither of whom does (one's a pathologist and one's a researcher) and who have to face patients in this very situation probably have a better grasp on what's going on than this. Believe me, it's almost as hard for the physician to face these patients week after week, as he throws drug after drug at the disease, slowing it temporarily, only to have it come roaring back. The problem, of course, is that cancer is very, very hard, and 36 years is not really that long a time when tackling such a difficult problem. For cancer, it's virtually all or nothing. Here's what I mean. Let's say your therapy kills 99% of tumor cells. That sounds pretty good, right? In many situations it would be good enough. But often not in cancer. The reason is that a tumor mass can have many billions of cells. Killing 99% of a small tumor mass with a billion cells still leaves 10 million cells behind. Moreover, cancer is a deadly and ever-changing foe (more on that later) that develops resistance after each round of therapy that's thrown at it. Given these facts, it's actually rather amazing that we can cure any cancer at all, at least when it comes to the hematopoietic malignancies, which can't be surgically cut out. (Indeed, one problem with the series is that Miklos and Baird seem to discuss only solid malignancies and basically completely ignore lymphomas and leukemias.) The article may have a point when it points out the problems with the "blockbuster drug" tendencies in big pharma, the profit motive, and how the media play into that phenomenon, but Miklos and Baird seem to have little clue about just how incredibly difficult a problem cancer represents. Worse, they even seem to fall into the same mentality that alties do by viewing "cancer" as a single disease, when it is in reality hundreds of different diseases.

In the next part of the series, Miklos and Baird make the mind-numbingly obvious observation that very few patients die of their primary cancer but instead die of metastases before scoffing at the increased survival since 1970 of patients with metastatic cancer:

The latest figures from the NCI's publicly accessible databases reveal the relative 5 year survival of patients with various metastatic cancers over the 30 year period, 1973 to 2004 (seer.cancer.gov):

survival with metastatic breast cancer improved from 19 to 23 percent.
survival with metastatic colorectal cancer improved from 6 to 9 percent.
survival with metastatic prostate cancer improved from 28 to 34 percent.
survival with metastatic lung cancer improved from 1 to 2 percent.

(Editor's note: these figures gathered over more than three decades amount to this: they can claim a "100% improvement" on lung cancer by improving survivability from 1% to 2%, and "a 50% increase in survival from colorectal cancer" breathlessly in press reports when they went from 6 to 9% improvement. This after multiple decades and tens of billions of dollars. That's how the gullible press, government-paid bureaucrats, and corporate spin-doctors actually express things. --Dean)

The improvements in survival are less than 0.2 percent per year, a miniscule change. Dr Jane Weeks, an oncologist at the Dana-Farber Cancer Center provides clinical candour; A surprisingly high proportion of patients with metastatic solid tumors don't realize that there is no chance for cure. I've wondered how many patients in exactly that situation have been shocked to learn otherwise from the coverage about Elizabeth Edwards (17).

The truth is that all metastatic cancers are incurable despite the enormous sums of money poured into research and drug development, as well as the large amounts of chemotherapy, radiation and new drugs that have been poured into patients (24,25).

Actually, Miklos and Baird make a bit of a mistake here. They average the improvements in survival over 35 years, when in fact, many of these advances have occurred over the last decade or two. Moreover, they're just plain wrong when they blanketly state that "all" metastatic cancers are incurable. That's a load of bullshit, plain and simple. It's true that few metastatic cancers are curable, but that's a different thing than saying that they are "all" incurable. For example, metastatic testicular cancer is potentially curable with chemotherapy and radiation. Metastatic sarcoma, if it's disease that can be completely resected, is also potentially curable with surgery. Ditto for the most commonly cited example of a solid tumor that's curable with surgery: Colorectal cancer metastatic to the liver. If the disease is confined to the liver and can be completely resected (admittedly a minority of patients), five year survival rates of up to 40% can be expected.

It's also funny that they use Elizabeth Edwards as an example. True, her metastatic breast cancer is incurable, but, as I discussed earlier, she can reasonably expect to have several years before her disease claims her. It may be true that the overall survival from metastatic breast cancer has only increased by around 20% (relatively speaking), but in certain subtypes there is hope. Worse, as I alluded to earlier, Miklos and Baird's conception apparently encompasses only solid organ tumors, ignoring, for example, lymphomas, which are often present in multiple lymph node basis but can still be cured with chemotherapy and radiation. For allegedly being "cancer researchers," these guys sure do seem pretty clueless about a number of basic clinical facts about cancer and oddly unaware of advances in the field. Most likely, it's because neither of them are clinicians, and neither of them appears to have designed or interpreted a therapeutic clinical trial. They also seem to discount the fact that, even if progress against cancer in terms of survival has been disappointing, we have been steadily moving towards treatments that are just as efficacious but much less morbid. This is not something to be dismissed or discounted. In the case of surgery, this means smaller operations. In the case of chemotherapy, and other treatments, less toxicity. And we have done well at that over 35 years. Back in Nixon's day, for example, the standard of care for even small breast cancers was radical or modified radical mastectomy, the former of which was a very disfiguring operation. Now around 60-80% of women with breast cancer can be treated with lumpectomy, and we don't routinely remove all the axillary lymph nodes anymore, greatly decreasing the complications of lymphedema. Many colon cancers are now treated with laparoscopic approaches.

I will admit, though, in this second installment, looking at treatments for metastatic cancer discussed at the American Society of Clinical Oncology Meeting (which, by the way, I attended), were rather disappointing. It is not unreasonable to ask if the marginal benefits of many of these drugs is worth their high cost. However, as you will see, Miklos and Baird's alternative is hardly the radical break with established methodology that they seem to be leading the reader to believe that it will be.

The next part is a bit annoying. It's not so much that Miklos and Baird are incorrect; it's more the way they present some very basic tumor biology. Basically, they point out that tumors are genetically diverse, with many different subpopulations containing different genetic derangements. Moreover, they point out that only a small proportion of cells from a primary tumor metastasize. They conclude that gene profiling is more or less useless and that this is The Way:

Current personalized cancer medicine focuses on single letter mutations rather than the massively disrupted DNA component of cancer. Half a century of genetics, however, shows that the effects of massive changes involving many genes dwarf the effects of single letter mutations. Analyses of additions and deletions of DNA in experimentally manipulable organisms reveal that varying the dosage of large chunks of DNA has far more important biological effects on the flexibility of genetic operating systems than the small scale mutational changes that can be induced in normal cells (44).

So how have we reached this preoccupation with personalized DNA profiling of mutations when our answer lies not in the bulk of the tumor, but in the tiny population of maverick cells with their massively disrupted DNA contents?

Thanks, guys--because we know that cancer researchers never would have thought to look at the genetic derangements in metastatic cells that lead them to metastasize as a means of figuring out how to combat metastasis. I guess those thousands of articles on the topic that can be found in MEDLINE are evidence of just how little interest is shown in the genetics of metastasis. There was actually a rather good recent review in The Lancet back in May. In essence, there are three main hypotheses of metastasis: the classical clonal selection concept, in which rare tumor cells spontaneously generate metastatic potential and then are selected for as they break off from the tumor mass; the "predetermined/hardwired" hypothesis, in which a metastasis gene expression pattern is (or is not) expressed in most cells in primary tumors, making some tumors "hard-wired" to metastasize; and the host predisposition hypothesis, in which host factors play the dominant role. In fact, depending upon the tumor, different combinations of these three hypotheses play a role in differing degrees. The interesting thing coming out of recent research is that metastatic potential may well be already determined even in small tumors. There is, for example, a test that looks at a panel of genes in early stage breast cancer, the OncoType DX. It identifies a gene expression signature that is associated with a higher risk of metastasis and allows oncologists to spare low risk patients from chemotherapy. Yet, Miklos and Baird almost completely poo-poo genetic testing of tumors for such purposes as "useless."

The next part of the article is where things get stupid. (I'm sorry, but there's simply no other word for it.) For example, Miklos and Baird state:

Recent population-based data on two famous breast "cancer" genes calls into question the clinical significance of the extensively studied mutations in these genes, for which many women are tested (84)...

Most women are aware of these two genes [BRCA1 and BRCA2] and dread that they may carry mutations within them. After decades of detailed examination of mutations in these genes, however, definitive clinical data now reveal that breast cancer specific rates of death among women who are carriers of BRCA1 or BRCA2 mutations, are no different to those of women without these mutations (84). Clinical data have thus rendered irrelevant a decade or more of mutation-based research data.

Uh, no. The above statement is breathtakingly dumb. Here is the article that Miklos and Baird cite to back up this rather odd statement:

Rennert G, Bisland-Naggan S, Barnett-Griness O, Bar-Joseph N, Zhang S, Rennert HS, Narod SA. Clinical outcomes of breast cancer in carriers of BRCA1 and BRCA2 mutations. N Engl J Med. 2007 Jul 12;357(2):175-6.

BACKGROUND: Some features of breast cancer in women with a BRCA1 mutation suggest that hereditary breast cancer has a poor outcome. We conducted a national population-based study of Israeli women to determine the influence, if any, of a BRCA1 or a BRCA2 mutation on the prognosis in breast cancer. METHODS: We obtained data on all incident cases of invasive breast cancer that were diagnosed from January 1, 1987, to December 31, 1988, and recorded in the Israel National Cancer Registry. We requested a paraffin-embedded tumor block or an unstained slide and the corresponding pathological and clinical records for all such cases. DNA extracted from the tumor specimens was analyzed for the three founder mutations in BRCA1 and BRCA2. For each subject, available pathological and oncologic records were reviewed. RESULTS: We were able to retrieve a pathological sample from 1794 of 2514 subjects (71%). Among those women, we obtained medical records for 1545 (86%). A BRCA1 or BRCA2 mutation was identified in 10% of the women who were of Ashkenazi Jewish ancestry. The adjusted hazard ratios for death from breast cancer were not significantly different among mutation carriers and noncarriers (hazard ratio among BRCA1 carriers, 0.76; 95% confidence interval [CI], 0.45 to 1.30; P=0.31; hazard ratio among BRCA2 carriers, 1.31; 95% CI, 0.80 to 2.15; P=0.28). Among women who were treated with chemotherapy, the hazard ratio for death among BRCA1 carriers was 0.48 (95% CI, 0.19 to 1.21; P=0.12). CONCLUSIONS: Breast cancer-specific rates of death among Israeli women are similar for carriers of a BRCA founder mutation and noncarriers.

It is fortunate for me (and unfortunate for Miklos and Baird) that I had just so happened to have presented this very paper at a recent meeting of our surgical oncology journal club. It's a so-so paper, a retrospective study, and it's certainly not "definitive" clinical evidence, but even if it were the greatest thing since sliced bread, it doesn't support Miklos and Baird's contention at all! What the paper concludes is indeed that carriers of BRCA mutations do more or less as well in terms of survival as noncarriers, but it should be pointed out that this is a comparison of two groups of women who all already have been diagnosed with invasive breast cancer. The concept that BRCA mutations predispose to cancer but that the cancers produced by them are, as far as we can tell, not more aggressive than sporadic cancers not due to hereditary BRCA mutations appears to be a simple concept that seems to have eluded Miklos and Baird. They seem to be saying that knowing a woman's BRCA mutation status is always useless, just because once a carrier has progressed to invasive cancer, the presence of this mutation may not predict how aggressive the cancer will be. Their overblown conclusion (that "clinical data have thus rendered irrelevant a decade or more of mutation-based research data") is completeley unjustified (or, as I've taken to saying about such idiocy, a fetid load of dingo's kidneys). In actuality, knowing a woman's BRCA mutation status is very important in at least one area; Genetic counseling. If a woman has a BRCA mutation predisposing to cancer and has daughters, she may very well want to tell them to be tested, so that they can consider either closer observation or risk reduction strategies.

So, after a very hit-and-miss affair with some reasonable criticisms of our cancer research establishment interspersed with stinkbombs of misleading or ignorant statements like the one about BRCA mutations scattered every so often throughout the series, we finally get to the conclusion. So, what is the great insight that Miklos and Baird have that will, if the NCI, big pharma, and the medical profession would only adopt it, change this "35 years of failure" that Miklos and Baird (and Dean Esmay) seem to think our research effort has yielded? Surely it must be brilliant! Surely it must be revolutionary! So, let's hear it:

Increased investment is needed in imaging and screening technologies for early detection such as nanotechnologies that are able to detect whether cancer cells have left the primary tumor and accumulated in lymph nodes (98). It is time for early diagnosis and prevention to take precedence over research aimed at shutting down cancers after they have spread (99). Cancer agencies, well meaning charitable organizations and foundations need to stop being automatic teller machines for basic science (100).

I should win a prize. I predicted it! I predicted that, at the end of this long series, The Answer to the Problem of Cancer Research in the U.S. would be: More early detection and more detection of microscopic metastases. Oh, and getting people to quit smoking:

One immediate preventative measure is staring everyone in the face. More people in the US will die from lung cancer in 2007 (approximately 160,000), than all the deaths from colorectal, breast, pancreas and prostate cancer combined. Tobacco usage is also a risk factor for the development of many other cancers. Stop its usage and the incidence of lung and other cancers will be greatly reduced.

Well, blow me over with a feather! We obviously corrupt cancer researchers, apparently blinded by equal parts hubris, NIH grants, and big pharma, never would have thought of either of those things--never in a million years! Get people to stop smoking? Who'd have thought of it? Detect cancer at an earlier stage? Briliiant! Truly, I am not worthy to comment on the glorious solution to the problem of cancer proposed here that apparently cancer researchers are unable to see!

There's just one problem.

Although it would indeed truly be great if we could eliminate smoking, which would certainly greatly reduce the rates of many cancers, not to mention the rate of cardiovascular disease as an added benefit (although I must point out that it would take a couple of decades to see the fruits of these efforts, even if all smokers kicked the habit tomorrow), Miklos and Baird's fetish for early detection über alles shows how out of touch they are with clinical research. Sure, early detection is very useful for many cancers. Mammography, for example, definitely saves lives. But it doesn't save as many as you would think, a 25% reduction in breast cancer mortality in women over 50 and a lesser benefit in women below 50. Miklos and Baird would no doubt retort that this is due to a limitation of the technology; partially it is, but it's also largely biology as well. There is a point of diminishing returns in early detection and intervention, mainly because not all early cancerous lesions go on to develop into life-threatening disease. Prostate cancer is the classic example, where 40% of men in their 60s and 75% of men in their 80s have histologic evidence of the disease, but only 1% go on to develop cancer. Moreover, prostate cancer is often slow growing, such that the majority of men who develop it die of something else (such as old age) before succumbing to the cancer. Excessively zealous early detection and intervention has the potential to do real harm in the cases of such cancers. I once again not-so-humbly suggest that Drs. Miklos and Baird go back and review concepts such as lead time bias, length bias, and the Will Rogers effect before making such blanket statements if they want to avoid looking so clueless in the future, and then read a bit about the controversies about the use of MRI as a more sensitive screening tool for breast cancer (a topic about which I'll probably have more to say later this week). Indeed, the lesson of using CT scans to screen even a high risk group (smokers) for lung cancer is a good cautionary tale about how early detection is no panacea. You'd think it would be a slam dunk for this sort of case, given that CT scans can detect small nodules better than chest X-ray, but it's still very unclear whether CT scanning saves lives. Recent studies suggest that it doesn't, while it does expose patients to more invasive biopsies.

There is always a price to be paid for ever earlier detection. Always. The question is always whether the benefits outweigh the risks, a fact to which nonclinicians like Miklos and Baird seem oblivious. They may envision amazing advances in technology that increase our ability to detect cancer to very early clumps of cancerous cells and our ability to detect microscopic metastases to the same degree. In fact, we can actually already do this to some extent for breast cancer and melanoma, by the way, using sentinel lymph node biopsy. This procedure brought about endless arguments over whether the finding of a small clump of metastatic breast cancer cells in a sentinel lymph node means anything for prognosis--the evidence thus far favors the contention that it probably does not. The problem, of course, is that the more sensitive the test, the more invasive interventions to remove or treat abnormalities that might never manifest themselves as clinical disease or, in the case of established cancers, that have no bearing on overall prognosis.

Don't get me wrong, better early detection for many cancers is generally a good thing. The problem is that we are probably starting to reach the point of diminishing returns for the screening of at least some common cancers, for example, prostate (and possibly breast) cancer, while early detection of certain deadly cancers like pancreatic cancer could potentially save many lives because only early, resectable tumors are even potentially curable with surgery. Miklos and Baird massively oversell early detection of cancer and microscopic metastases and completely ignore the not insignificant downside to detecting cancer at ever earlier points in its development, points when it is impossible, at least as of yet, to predict which cancerous lesion will progress into life-threatening tumors. They may have a point that it could be advisable to shift some funding and emphasis to early detection, but it's important to have a realistic idea of what early detection can and cannot do for cancer. Miklos and Baird clearly do not. Moreover, this is not at all a revolutionary or even a particularly creative suggestion. Sadly, if this series were a fireworks display, the beginning would be a bunch of Ladyfingers, loud and showy but weak, the middle a few lone bottle rockets interspersed with a few stinkbombs, and the final installment a single, sputtering sparkler fizzling out.